Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A conceptual and empirical study of issues in stereotype research : some reflections on the definition… McTiernan, Timothy John 1977

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1977_A8 M32_8.pdf [ 11.15MB ]
Metadata
JSON: 831-1.0094149.json
JSON-LD: 831-1.0094149-ld.json
RDF/XML (Pretty): 831-1.0094149-rdf.xml
RDF/JSON: 831-1.0094149-rdf.json
Turtle: 831-1.0094149-turtle.txt
N-Triples: 831-1.0094149-rdf-ntriples.txt
Original Record: 831-1.0094149-source.json
Full Text
831-1.0094149-fulltext.txt
Citation
831-1.0094149.ris

Full Text

A CONCEPTUAL AND EMPIRICAL STUDY OF ISSUES IN STEREOTYPE RESEARCH: SOME REFLECTIONS ON THE DEFINITION AND MEASUREMENT OF STEREOTYPES, AND AN EXPERIMENT ON RECIPROCITY AND STEREOTYPE FORMATION by Timothy John McTiernan B.A. (Mod.), Trinity College, University of Dublin, 1973 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES Department of Psychology We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA April, 1977 © Timothy John McTiernan, 1977 In p r e s e n t i n g t h i s t h e s i s in p a r t i a l f u l f i l m e n t o f the r e q u i r e m e n t s f o r an advanced degree at the U n i v e r s i t y o f B r i t i s h Co lumb ia , I a g ree that the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s tudy . I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y purposes may be g r a n t e d by the Head o f my Department o r by h i s r e p r e s e n t a t i v e s . It i s u n d e r s t o o d tha t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l not be a l l o w e d w i thout my w r i t t e n p e r m i s s i o n . Department o f f W c ^ t > L o & ^ The U n i v e r s i t y o f B r i t i s h Co lumbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date l^qj^J tlH i i • ABSTRACT The present study includes a c r i t i c a l evaluation of the t r a d i t i o n a l stereotype l i t e r a t u r e as well as an argument i n support of a broader d e f i n i t i o n of the term. I t contains a discussion of the appropriateness of the a d j e c t i v a l check l i s t and the open-ended response format as research instruments. A preliminary model of the stereotype formation process i s outlined and the method, r e s u l t s , and conclusions of a stereotype formation study are presented. The commonly espoused d e f i n i t i o n s of stereotypes as being e i t h e r overgeneralizations, consensual b e l i e f s , or r i g i d and i r r a t i o n a l substitutes for thought were shown to be l o g i c a l l y weak and to s u f f e r from many of the l i m i t a t i o n s inherent i n Lippmann's (1922) o r i g i n a l conceptualization. A broader perspective was advocated i n which stereotypes were defined as being s o c i a l concepts which d i f f e r from other concepts i n that they can include personality terms as w e l l as terms r e f e r r i n g to non-social, p h y s i c a l c h a r a c t e r i s t i c s . As such, stereotypes were considered to be concepts about groups and i n d i v i d u a l s , the performers i n the s o c i a l environment. In discussing the nature of stereotypes, the d i s t i n c t i o n was made between personal stereotypes, the concepts i n our heads, and c u l t u r a l and s o c i a l stereotypes, which are s o c i o - c u l t u r a l rather than psychological phenomena. A d d i t i o n a l l y , i t was argued that, although they are the most frequently reported data, check l i s t c haracterizations may not be appropriate i i i analogues to the spontaneously produced personal stereotypes that are used by i n d i v i d u a l s i n the course of t h e i r d a i l y l i v e s . The core components of these personal stereotypes, i n that they are based on r e c a l l , are better measured by unstructured, open-ended instruments. Check l i s t s , on the other hand, serve as indexes of those terms which we are prepared to iendorse a f t e r they have been brought to our a t t e n t i o n i n the course of conversations and s o c i a l a c t i v i t i e s . Consequently, check l i s t s , because they are based on recognition rather than r e c a l l , provide a sample of the peripheral elements of stereotype c h a r a c t e r i z a t i o n s . The point was made that i n order to obtain a comprehensive understanding of the contents and structure of stereotypes i t i s necessary to use a combination of an open-ended and a modified check l i s t procedure when conducting research. Further conceptual analysis indicated that a comprehensive account of the stereotype formation process requires at l e a s t two models, one def i n i n g the variables which govern the development of stereotypes about targets possessing c l e a r c u l t u r a l d e f i n i t i o n s , and the other de l i n e a t i n g the form of i n d i v i d u a l s ' more general "stereotype formation s t r a t e g i e s . " The second part of t h i s thesis focusses on a preliminary, three-step model of the general, rule-governed stereotype formation process. The model postulates that : i n d i v i d u a l s use a v a i l a b l e information to construct demographic p r o f i l e s of c u l t u r a l l y undefined targets; a t t r i b u t e p a r t i c u l a r "world views" and value systems to the targets on' the basis of the i n f e r r e d demographic c h a r a c t e r i s t i c s ; and then make t r a i t a s c r i p t i o n s on the basis of t h e i r inferences about the targets' general perspectives, using reference group norms as t h e i r judgemental iv anchors in the process. The influence of "world view" variables on the stereotype formation process was examined in an experiment which employed information about a target's characterization of the respondents' salient reference group as an independent variable. Three versions of an audiovisual documentary about the Orkney Islanders were constructed and shown to groups of Canadian sea cadets. Sixty-seven respondents saw a control film containing a general description of the Islanders' way of l i f e , sixty-eight saw a modified version of the film which included "examples" of the Islanders' "highly unfavourable stereotypes of Canadians," and fifty-eight were presented with a positive reciprocal script in which the Islanders were described as characterizing Canadians very favourably. The film scripts were balance with respect to length, organization, and the polarity of the reciprocal stereotype information. A post-film questionnaire was administered, and free response and check l i s t instruments were used to measure the respondents' newly formed stereotypes one week after the presentation of the stimulus materials. The findings determined that the respondents made inferences about an unfamiliar target group's stereotypes of their own national group in the absence of definite information about the nature of these "reciprocal stereotypes;" that reciprocal stereotype information was attended to when i t was included in a documentary script about the Orkney Islanders; and that the characteristics of the personal stereotypes varied (1) as a function of the type of reciprocal stereotype information presented in the stimulus materials, and (2) in ( V relation to the manner in which this reciprocal stereotype information was construed by the respondents. The characterizations of the positive reciprocal group members were more favourable and more differentiated than those of the negative reciprocal respondents. Moreover, there was more agreement among the individuals in the positive reciprocal condition than among the members of the negative reciprocal group concerning which particular traits could be considered to be characteristic of the Orkney Islanders. The stereotypes of the non-reciprocal respondents f e l l between those of the positive and negative reciprocal group members on a l l of the measured dimensions. The implications of these results were discussed and refinements to the general stereotype formation model were suggested. v i TABLE OF CONTENTS Page Introduction . 1 PART ONE - ON THE DEFINITION AND MEASUREMENT OF STEREOTYPES 3 CHAPTER ONE STEREOTYPES: A FURTHER LOOK AT LIPPMANN'S INFLUENCE ON THE FIELD AND A CONCEPTUAL REORIENTATION 4 1. Stereotypes and the Individual 4 Traditional perspectives 4 Some criticisms of traditional assumptions 6 Llppmann's discussion of stereotypes 10 The context 10 The pseudo-environment and stereotypes. . 11 Stereotypes as "social" concepts 15 Implications of the "social concept" definition of stereotypes 17 2. Stereotypes and social groups 19 Cultural stereotypes 19 Social stereotypes 21 Cultural and social stereotypes as context 25 Summary 27 CHAPTER TWO MEASURING INSTRUMENTS AND STEREOTYPE RESEARCH . . . 29 The adjectival check l i s t : Some problems 29 The appropriateness of stereotype measures 30 Relevant research findings 32 Core and peripheral stereotype components and their measurement 33 v i i Page Implications of a revised measuring procedure. . . . . . . . . . . . . . . . 34 Summary . . . . . . . . . . . . . . . . . 36 PART TWO - KNOWLEDGE OF THE TARGET'S CHARACTERIZATION OF THE RESPONDENTS * NATIONAL GROUP AND THE STEREOTYPE FORMATION PROCESS . . . 38 CHAPTER THREE A PRELIMINARY HYPOTHESIS CONCERNING RECIPROCITY AND STEREOTYPE FORMATION . . 39 Stereotype formation and the social context . . . . 39 Psychological processes and the development of stereotypes . 40 A preliminary model of the stereotype formation process 42 The research problem 44 Summary . . . . 47 CHAPTER FOUR METHOD 49 Overview 49 Stimulus materials 52 1. The film-tape sequence 52 a. The film 52 b. Primary script 53 c. Variations 54 d. Presentation 56 2. Target groups 57 Questionnaire materials 58 1. Post-film questionnaire 58 2. Stereotype questionnaires 59 a. Open-ended format 59 b. Check l i s t format 59 c. Information forms and other questionnaires 60 v i i i Page Respondents 61 Setting 65 Procedure 66 Phase 1 66 Phase 2 69 Special problems associated with the data c o l l e c t i o n 71 Scoring procedure and methods of analysis . . . . . 74 1. S o c i a l stereotypes 74 2. Personal stereotype variables 75 3. Primary data analyses 77 Review 79 CHAPTER FIVE PERESUL'TSKANDIDISGUSSION 80 1. Manipulation checks and the treatment of methodological problems 80 Manipulation checks 80 Pretest on stereotype data from the negative r e c i p r o c a l group 85 Homogeneity of variancespretests 86 Correlations between age and stereotype v a r i a b l e s . 88 Summary of methodological r e s u l t s 88 2. Presentation and discu s s i o n of experimental r e s u l t s . . . 90 a. P o s t - f i l m impressions of the Orkney Islanders and estimates of the Islanders' r e c i p r o c a l stereotypes estimates of thaofsCanadians 90 Post - f i l m impressions of the Orkney Islanders . . . 90 Respondents' estimates of the Orkney Islanders' stereotypes of Canadians 97 ix Page Contents 97 Structure . . . . . . . . . . . . . . . . 99 Overview . . . . . . . . . . . . . . . . . . . . . . 102 b. The contents and structure of the respondents'. personal stereotypes of the Orkney Islanders 102 Contents 102 Structural attributes 110 Differentiation and favourableness . . . . 110 Differences between groups 114 Differences between,stereotype components ' 114 Differences within stereotype components 116 Respondents' use of stimulus material information . . 130 Overlap between the content of the respondents' stereotypes of the Orkney Islanders and their concepts of Canadians 132 Review of the analyses of the personal stereotype variables 135 c. Canonical correlation analyses of the relationship between the personal stereotype variables and the estimated reciprocal stereotype measures 136 d. Review of the findings 142 Chapter overview 142 CHAPTER SIX SUMMARY AND CONCLUSIONS 143 PART THREE - OVERVIEW AND FINAL COMMENTS . . . . 153 CHAPTER SEVEN OVERVIEW AND FINAL COMMENTS 154 FOOTNOTES 156 BIBLIOGRAPHY 164 APPENDIXES 179 X Page APPENDIX A SCRIPTS FOR THE TAPE RECORDED DOCUMENTARY PRESENTATION . . . . . . . . . . . . . . 180 Common portion of s c r i p t . . . . . . . . . 182 1. Non-reciprocal condition . . . . . . . . . . . . 192 2. P o s i t i v e r e c i p r o c a l stereotype s c r i p t 194 3. Negative r e c i p r o c a l stereotype s c r i p t 197 Concluding remarks (used i n a l l three s c r i p t s ) 200 Quotation sources 201 APPENDIX B PROCEDURE FOR SELECTING THE TRAIT TERMS USED IN THE STIMULUS MATERIALS 202 1 203 Procedure 203 Results 204 2 . 205 Procedure 205 Results 205 Note 206 APPENDIX C PROCEDURE FOR SELECTING THE ADDITIONAL TARGET GROUPS 207 Method „.„.,.>.„...„.*.... 208 Respondents 208 Procedure " 208 Scoring procedure 209 Results 209 Discussion 211 x i APPENDIX D POST-FILM QUESTIONNAIRE . . . . . . . . . . . . . . . 214 APPENDIX E EXAMPLE OF THE OPEN-ENDED STEREOTYPING FORMAT 218 APPENDIX F EXAMPLE OF THE CHECK LIST MEASURING PROCEDURE 220 APPENDIX G PERSONAL INFORMATION QUESTIONNAIRES USED IN STEP 1 AND STEP 2 OF THE EXPERIMENT . . . . . . . 224 APPENDIX H TABULATION SHEET FOR NOTE-TAKING DURING FILM SEQUENCE 227 APPENDIX I "SOURCES OF STEREOTYPE INFORMATION" QUESTIONNAIRE . 229 APPENDIX J BASIC RIGHTS AND PRIVILEGES OF VOLUNTEER SUBJECTS 231 APPENDIX K VERBAL DIRECTIONS TO RESPONDENTS BEFORE AND AFTER FILM PRESENTATION 233 APPENDIX L INTRODUCTORY PASSAGE FOR PHASE TWO OF PROCEDURE 238 APPENDIX M VERBAL DIRECTIONS TO RESPONDENTS DURING PHASE TWO OF PROCEDURE 240 Step 1 241 Step 2 245 Notes 247 APPENDIX N EXAMPLE OF CHECK LIST RESPONSES 248 APPENDIX 0 TABLE OF RESPONSES TO QUESTIONS CONCERNING THE RESPONDENTS* POST-FILM IMPRESSIONS OF THE ORKNEY ISLANDERS 251 APPENDIX P TABLES OF MEANS AND STANDARD DEVIATIONS OF SCORES ON THE DIMENSIONS OF THE ESTIMATED RECIPROCAL ORKNEY ISLANDER STEREOTYPES OF CANADIANS 257 x i i LIST OF TABLES Page 1. The t r a i t s used to describe the Orkney Islanders i n the stimulus materials and the terms used to represent the Islanders' p o s i t i v e and negative c h a r a c t e r i z a t i o n of Canadians ., 55 2. The l o c a t i o n of permanent residence of the respondents i n each of the experimental conditions and i n the t o t a l sample 62 3. The age d i s t r i b u t i o n of respondents i n each of the three experimental conditions and i n the t o t a l sample . . . . 63 4. The number and percentage of the i n d i v i d u a l s i n each of the experimental groups and i n the t o t a l sample choosing each of the possible responses to the post-f i l m questions pertaining to the experimental manipulations 81 5. Modal responses of p o s i t i v e , negative, and non-reciprocal group members to questions concerning t h e i r p o s t - f i l m impressions of the target 91 6. Most frequently occurring items i n the respondents' estimates of the Orkney Islanders' stereotypes of Canadians 93 7. Means and standard deviations f o r the d i f f e r e n t i a t i o n and favourableness scores f o r the core and peripheral components of the respondents' estimates of the Orkney Islanders' r e c i p r o c a l stereotypes of Canadians 98 8. S o c i a l stereotypes of the Orkney Islanders constructed from the most frequently occurring items i n the personal stereotypes of the respondents i n the three experimental conditions 103 9. Mean commonness and "'percentage of common terms' scores f o r the core and peripheral components of the personal stereotypes of the Orkney Islanders 107 x i i i Page 10. Analyses of variance of the commonness and 'percentage of common terms' scores 108 11. Analyses of variance of the differentiation and favourableness scores 112 12. Mean differentiation and favourableness scores for the core and peripheral components of the personal stereotypes of the Orkney Islanders 113 13. Mean number and percentage of positive, neutral, and negative words In the core and peripheral components of the group members' characterizations of the Orkney Islanders 1117 14. Analyses of variance of the numbers and percentages of favourable, neutral, and unfavourable words in the respondents' characterizations 119 15. Canonical correlations of personal stereotype and est:'^estimated reciprocal stereotype variables for the non-reciprocal group - negative variables excluded . . . . 139 xiy "LIST 'OF FIGURES 1. Design structure of the two- and three-way analyses of variance 78b 2. Mean number of favourable, n e u t r a l , and unfavourable words i n the core and periphery of the personal stereotypes of the members of each of the three experimental groups . . . 115 3. Mean percentages of favourable, n e u t r a l , and i unfavourable terms i n the personal stereotypes of the p o s i t i v e , negative, and non-reciprocal group members . , 122 4. Mean percentages of favourable, n e u t r a l , and unfavourable items i n the core and peripheral components of the respondents'personal jestereotypes 123 5. Schematic diagram of the three-step stereotype formation model 150 XV - ACKNOWLEDGEMENT I am indebted to a great many people for t h e i r invaluable assistance and encouragement i n the undertaking and completion of t h i s t h e s i s . Foremost among these are: Mr. James Gove, whose many e f f o r t s made i t possible f o r ' me to obtain my respondents; Dr. Robert E. Knox, who gave me a l l the rope I needed, who waited p a t i e n t l y f or two years, and who provided constant moral and i n t e l l e c t u a l support throughout the period i n which I worked on t h i s study; Miriam, my wife, who i s the only person a l i v e who knows the text by heart and who, despite that f a c t , has given me the emotional encouragement and the helping hand needed to complete the enterprise; and Leah, my daughter, who,maintained a very p h i l o s o p h i c a l approach to my regular disappearances at weekends. In addition, I very much appreciate the c r i t i c i s m s and thoughtful suggestions of Drs. Tom Storm and Ralph Hakstian. Keith Humphrey and Bob Boutdlier helped immensely i n the preparation and administration of the stimulus materials, and Dr. Val Hunt and Kathleen Sun spent many hours respectivelyutyping the thesis proposal and scoring the responses. I am extremely g r a t e f u l for t h e i r f riendship. I would l i k e also to express my thanks to Gordon Handford, Rod Borrie, Pat Vining, Werner Neufeld, and Georgia Shelton who acted as judges i n the p i l o t stages of the project, to Peter Van Oot and David Cox f o r providing access to p i l o t subjects, to Jackie Humphrey f o r c o l l a t i n g questionnaires, to Eddie Erikson for r o s t e r i n g some data, and to John xv i Simmons for h i s t e c h n i c a l assistance. I am e s p e c i a l l y g r a t e f u l to the cadets from the National Sea Cadet Camp at H.M.C.S. Quadra who were f r i e n d l y and w i l l i n g respondents Intthe experiment, and to Commander W.R. Vipond CD(CZL/S), Sub-Lieutenant B.G. Kleeberger (CIL/S), Lieutenant Commander D. L e i t c h (CIL/S), and Lieutenant _ W. Abbott (USNSCC) whose w i l l i n g cooperation and o r g a n i z a t i o n a l J G a b i l i t i e s made my stay at Quadra a rewarding and enjoyable one. P a r t i c u l a r thanks are due also to Mr. John Steele, to Lieutenant L.G. Jenks of C.F.B. Comox, to Lieutenant Colonel J.L. Frazer. Captain Ti s d a l e and Lieutenant P o i r r i e r of C.F.B. Chilliwack, and to the o f f i c e r candidates from the O f f i c e r Candidate Training School, C.F.B. Chilliwack f o r t h e i r help and cooperation during the I n i t i a l stages of the study. The experiment reported i n th i s t hesis was funded by a P r o v i n c i a l Government Careers 75 Grant, No. 1788, and by research support provided by Dr. R.E. Knox. I am indebted to Tina Chow, and p a r t i c u l a r l y indebted to V i r g i n i a Lee for t h e i r expertise and e f f i c i e n c y i n typing the major p o r t i o n of th i s manuscript. F i n a l l y , I would l i k e to acknowledge two people to whom I owe a great d e a l : my l a t e father, John McTiernan, who always supported me i n what I wanted to do, and Barbara Marston who was a dear and g i f t e d f r i e n d and who always found the time to be Interested i n other people's work. 1 A CONCEPTUAL AND EMPIRICAL STUDY OF ISSUES IN STEREOTYPE RESEARCH: SOME REFLECTIONS ON THE DEFINITION AND MEASUREMENT OF STEREOTYPES, AND AN EXPERIMENT ON RECIPROCITY AND STEREOTYPE FORMATION For the past f i f t y years the study of stereotypes has been a major topic i n the f i e l d of social psychology. Yet during that time l i t t l e hypothesis testing research has been produced and the amount of information gathered concerning the psychological and sociological variables affecting the formation and maintenance of stereotypes has been minimal. Rather, the majority of the studies have been descriptive in nature and, by and large, researchers have restricted themselves to defining those sets of charac-te r i s t i c s which the members of particular sample groups most frequently attribute to various national, socio-economic, and ethnic targets. The almost total concern with descriptive rather than hypothesis testing research is due, in part, to the lack of adherence to a general and clearly defined theoretical orientation. It is due, also, to a widespread and uncritical adoption of the adjectival check l i s t as a research instru-ment. The present study deals with some of the issues related to the definition and measurement of stereotypes and, in addition, i t reports an experiment which tested a preliminary hypothesis concerning the role played by reciprocity in the stereotype formation process. 2 This thesis w i l l be organized into three parts. The f i r s t part w i l l include a discussion of the limitations of the existing theoretical perspectives in the f i e l d , as well as an argument i n support of a broad definition of stereotypes (Chapter 1). It w i l l also contain a discussion of the appropriateness of the adjectival check l i s t as a research instrument (Chapter 2). In the second part, the conceptual basis for the stereotype formation experiment w i l l be presented (Chapter 3), and the method (Chapter 4), results (Chapter 5), and implications (Chapter 6) of the study w i l l be outlined. The f i n a l part of the thesis w i l l contain a brief overview of the conceptual and empirical sections (Chapter 7). 0 3 P A R T O N E ON THE DEFINITION AND MEASUREMENT OF STEREOTYPES 4 CHAPTER ONE; STEREOTYPES; A FURTHER LOOK AT LIPPMANN'S  INFLUENCE ON THE FIELD AND A CONCEPTUAL  REORIENTATION 1. Stereotypes and the Individual  Traditional Perspectives The empirical framework provided by the pioneering studies of Katz and Braly (1933, 1935) has served to highlight the central features of Lippmann's psycho-sociological treatise on stereotypes and has functioned to define these notions as important topics of interest within the f i e l d of prejudice research. Following i n the Katz and Braly tradition, a large body of social scientists has asked numerous groups of respondents to select and star the five or ten words in a l i s t of seventy or more that they considered to be most characteristic of each of a set of ethnic, occupational, or socio-economic targets. Other researchers have modified the basic procedure to some extent, but, i n a l l cases, when the responses have been tabulated the investigators have typically found that their subjects displayed a high level of agreement in the ascription of specific, unfavourable characteristics to particular minority r a c i a l groups and to unfamiliar, unpopular, and hostile nations (e.g., Dudycha, 1942; La Piere, 1936; Simmons, 1961). For example, in the i n i t i a l Katz and Braly study one hundred Princeton undergraduates labelled Negroes as being superstitious, lazy, happy-go-lucky, and ignorant, among other derogatory t r a i t s . With the notable 5 exception of studies by Karlins, Coffman, and Walters (1969) and Maykovich (1971, 1972), this pattern of ascriptions has continued to re-eaerge i n the vast body of research that focuses on stereotypes toward American Blacks (Gilbert, 1951; Hartsough and Fontana, 1970; Sigall and Page, 1971). Indeed, the characterization i s a general one and i s not subscribed to solely by white Anglo-Saxon Americans. There i s evidence to demonstrate that, in the past at least, some Blacks have accepted this image of them-selves (Bayton, 1941; Bayton and Byoune, 1947), while respondents in India, Lebanon, Great Britain, Pakistan, Taiwan, Canada, and Australia have held similar negative stereotypes (Hicks, Goldman, and Kang, 1968; Prothro and Melikian, 1954; Signori and Butt, 1972; Stanley, 1969). The negative content of such characterizations has caused many re-searchers and textbook writers to endorse the proposition that stereotypes act merely to justify faulty perceptions, unrecognized h o s t i l i t y , dogmatism, and ethnocentrism (Katz and Braly, 1933, 1935; Saenger and Flowerman, 1954; Simpson and Yinger, 1965). As Brown (1958) and Brigham (1971) have stated succinctly, the predominant mood of the f i e l d i s one which conveys the strong impression that stereotyping is wrong. They [the researchers] think that i t i s at least i r r a t i o n a l and probably wicked to subscribe to them (Brown, 1958, p. 364) However, while moral indignation continues to be expressed over the prevalence of stereotyping, there seems to be l i t t l e overall agreement concerning the precise nature of stereotypes, or concerning those features of stereotypes that are of the greatest theoretical significance. Most 6 of the speculation and commentary that has taken place in the f i e l d , however, f a l l s more or less within the confines of three major research trends. One central claim has always been that stereotypes are overgenerali-zations, which may be either invalid or ill-defined, but which are, in any case, unjustifiable according to some criterion (see Brigham, 1971). Supplementing this orientation is a perspective which considers stereotypes primarily to be consensual beliefs that are shared i n common by the majority of the members of a particular group (e.g., Gardner, 1973). A third viewpoint, which also has received substantial support, blends the former notions into a more extreme synthesis. In terms of this latter position, stereotypes are depicted as being rigid and irrational substitutes for thought which are borrowed by a great many people from "cultural mis-information" and "traditional nonsense" (Hayakawa, 1950). Although each has been widely subscribed to, these theoretical orientations have, in turn, been subjected to criticisms which have served to define their basic shortcomings and inadequacies (e.g., Brown, 1965; Fishman, 1956; Gardner, Rodensky, and Kirby, 1970). Some criticisms of traditional assumptions The question concerning the overgeneralizability of stereotypes is a d i f f i c u l t one to approach, the moreso because i t can revolve around either of two very different and unclearly defined assumptions. On the one hand, there is a presupposition that well-developed c r i t e r i a are available against which to evaluate the j u s t i f i a b i l i t y or 7 unj u s t i f l a b i l i t y of the contents of each particular stereotype. More often than not, however, i t i s d i f f i c u l t to generate the required c r i t e r i a . Additionally, there is a suspicion among some researchers (cf. Campbell, 1967, for a good summary) that when people attribute a charac-t e r i s t i c to a group as a whole (given, for example, that we are considering a stereotype of a particular national, regional, or occupational group) they may also automatically ascribe the characteristic to every individual within that group. If this i s the case, then indeed any stereotype would be an overgeneralization because, obviously, not every group member shares the same set of traits and personal qualities. However, stereotypes are not necessarily used i n this way. A stereotype of any particular group is a collective concept. It refers to general group characteristics in a manner that goes beyond basic individual differences. Cultural anthropologists have argued for long that this form of discourse is i n order. Consequently, the claim that stereotypes per se are overgeneralizations i s a non-sequitur un t i l at least two conditions have been met. >The f i r s t of these involves the development of appropriate c r i t e r i a to indicate that in broad terms any selected stereotype is an overgeneralization. The second condition requires a conclusive demonstration that most people, i f not a l l , commit the logical error of using class-specific stereotypes as unqualified descriptions of the many individuals with that class. The debate surrounding these issues is a complex one and a satis-factory solution to i t i s elusive. For example, Brigham's (1971) attempt to c l a r i f y the "overgeneralization claim" and to make i t more theoretically 8 acceptable has, Itself, been the target of substantial criticism by Gardner and his colleagues (Gardner, Rodensky, and Kirby, 1970). The proposition that stereotypes are consensual beliefs i s an appealing one. Supporting examples can readily be pointed to. Nevertheless, the closer one approaches the concept the more nebulous i t becomes, at least from a psychological standpoint. To say that a belief i s consensual i s to say merely that the believer shares i n the general agreement between the members of his reference group. It is not to make any comment about the psychological processes underlying the particular belief, nor about the relationship of that belief to others within the individual's cognitive system. Indeed, i t follows from the proposition at hand that a person's belief may be a stereotype from a North American point of view, for example, but an odd and idiosyncratic concept from a European perspective. Thus, in order to consider the notion of consensus as being fundamental to a definition of stereotypes one must adopt a sociological rather than a psychological focus and consider stereotypes only in relation to specific groups i n specific contexts. This severely restricts the generalizability of conclusions concerning both the process of stereotyping and the nature of stereotypes. Such limitations would seem to add to rather than a l l e -viate the conceptual confusion that now threatens the f i e l d . The further issue, concerning the supposedly i r r a t i o n a l , and r i g i d and inflexible nature of stereotypes i s , perhaps, the one that is most easily dealt with. In his early and comprehensive c r i t i c a l scrutiny of the literature, Fishman (1956) provided what i s s t i l l the best cautionary argument against the unquestioned acceptance of the r i g i d i t y thesis. The 9 general implication of this perspective i s that stereotypes remain fixed and completely unchanged over time. However: In any study of whether B's views of A remain the same over time, we must consider (a) whether the information reaching B concerning A has remained the same or not, (b) whether B (his needs, motives, interests) has remained the same or not, (c) whether A has remained the same or not, and (d) whether the interaction of B with A has remained the same or not. (Fishman, 1956, p. 37) Fishman contended that i t i s only when at least one of these factors has changed, without a related change in B's outlook toward A, that one can properly infer stereotype r i g i d i t y . Yet, social changes due to war (Haque, 1968, 1969; Meenes, 1943; Sinha and Upadhyay, 1960a, 1960b), and even the shore furlough of the Seventh Fleet of the U.S. Navy (Prothro and Melikian, 1955), have induced alterations i n the stereotypes of different samples of respondents. Furthermore, Gilbert (1951) and Karlins, Coffman, and Walters (1969), replicating the early Katz and Braly (1933) study with new generations of Princeton undergraduates, have noted changes in the content of stereo-types over a timespan of thirty-five years. They attributed such changes to shifting social patterns of interaction and to osc i l l a t i n g value systems. These findings effectively counter the proposition that a l l stereotypes are invariably r i g i d . They also reinforce Fishman's point that i n any longitudinal study "we must be careful to distinguish between constancy of stereotyping and ri g i d i t y of stereotypes (1956, p. 38)." As one considers the various criticisms that can be directed against each of the three theoretical-definitional perspectives the question arises 10 as to why, despite the weight of contradictory arguments and research findings, these viewpoints continue to be adhered to. The answer l i e s partially in the examples of widespread prejudice that are to be found in the early stereotype studies of Katz and Braly (1933, 1935), Bogardus (1950), and La Piere (1936). It can be found also in the direction provided by Lippmann's (1922) original and commonly cited speculations on the nature of stereotypes. Lippmann's discussion of stereotypes The context Immediately after the f i r s t World War a number of social commentators, including Walter Lippmann, published detailed analyses of twentieth century western l i f e s t y l e s . For his part, Lippmann adopted a Platonic stance. He used a "neo-Res Publica" model of the state as a framework for his analysis of post-war social structures. His conclusions were outlined in the many-faceted thesis that is contained within the pages of Public Opinion, f i r s t published in 1922. Not surprisingly, Lippmann f e l t that a combination of overwhelming technological development and role specialization was characteristic of an urban and industrial way of l i f e . More importantly, however, he argued that individuals liv i n g in such a milieu could no longer hope to maintain a general, inclusive focus on current events. The danger, according to Lippmann, is that the loss of a broad perspective f a c i l i t a t e s the misrepre-sentation of reality and leads to the distortion of one's own perceptions, goals, needs, and intentions. He identified these errors as being the unrecognized sources of contemporary s t r i f e , suffering, and bitterness. 11 In o u t l i n i n g the problems of modern l i f e Lippmann also prescribed a remedy. He urged the establishment of groups of s p e c i a l i s t s who could act as advisors, guides, and resource persons for those people who formulate and implement public p o l i c i e s . These s o c i a l i n t e r p r e t e r s were to be no mere c i v i l servants. Rather, from Lippmann's d e s c r i p t i o n they were supposed to be an e l i t e corps of i n t e l l e c t u a l s possessing most of the c h a r a c t e r i s t i c s of Plato's 'philosopher-guardians.' The pseudo-environment and stereotypes Lippmann approached the question of the misrepresentation of r e a l i t y and the d i s t o r t i o n of perceptions by fusing Plato's doctrine of i d e a l forms with Berenson's (1897) conception about the r o l e a r t plays i n providing us with 'stereotyped' features through which we can make sense of the world. He suggested that each i n d i v i d u a l possesses a pseudo-environment which acts as a buffer between the cor e - s e l f * and the outside world. This pseudo-environment can be likened to the Platonic world of form i n that i t embodies a fi x e d and i d e a l representation of the p h y s i c a l world of ever-changing mundane appearances. As such, i t has the i n t e r p r e t i v e function of unscrambling and f i l t e r i n g the unstable overload of information that comes from the environment at large. The structure of t h i s pseudo-environment was represented as a serie s of set, standard, p i c t u r e images, or stereotypes. These stereotypes supposedly act as f a m i l i a r forms through which new occurrences and novel s t i m u l i are assimilated, accommodated, and given meaning. The course of Lippmann's discussion about the nature and function of stereotypes was determined l a r g e l y by the framework of h i s broader s o c i o -12 p o l i t i c a l thesis, a framework which incorporates a conceptual reorientation of considerable significance. On a sociological level, Lippmann argued that p o l i t i c a l and journal-i s t i c censorship, the individual's occupational and social roles, and a host of other environmental factors, a l l serve to limit the amount, type, and accuracy of the information he obtains about his social world. Since such social distortion can result in misunderstandings and bad judgement, he called for the establishment of panels of experts ( p o l i t i c a l scientists) to better control the veracity and flow of information. In other words, he believed that a body of 'interpreters,' ' quality controllers,' and 'repairmen' was needed to rectify the existing faulty patterns of  communication. Building upon his argument, however, Lippmann also paid attention to the way i n which individuals process the information they receive from the world at large. As he did so he seemed to ignore the distorting features which he f e l t to be inherent i n the social environment. Rather, he presented the environmental milieu in relatively neutral terms as being merely "too big, too complex, and too fleeting for direct acquaintance (Public Opinion, 1961 ed., p. 16)." In contrast he stressed the point that individuals are inadequately equipped to deal with the variety of subtleties which surround them. Lippmann considered that each person's interpretation of the world, through the medium of his or her pseudo-environment, could be nothing but a simplified version of the actual state of a f f a i r s . He went on to surmise that "at the level of social l i f e , what is called the adjustment of man to his environment takes place through the 13 medium of fictions (1961 ed., p. 15).." Fictions which, can range " a l l the way from complete hallucination to the scientists' perfectly self-conscious use of a schematic model (1961 ed., p. 16)," but which are, i n a l l cases, reconstructions of the environment. There were at least two important consequences to such a conclusion. By asserting that individuals themselves can misrepresent and over-simplify their social reality Lippmann strengthened his argument in favour of the establishment of advisory panels of experts. At the same time, his particular focus ensured that his further discussion of stereotypes would emphasize their more negative features. Indeed, in terms of the general framework of the essay, he treated stereotypes as the antitheses of the informed and accurate perceptions of his ' p o l i t i c a l scientist advisors.' According to Lippmann, then, the stereotypes within our pseudo-environments determine that the evidence which infringes upon our senses i s "subject to illusions of defence, prestige, morality, space, time, and sampling (1961 ed., p. 154)." In other words, our reality testing i s inadequate. He argued, further, that because stereotypes are so well set, and because the underlying environmental reality i s i n such a constant state of flux, "there i s always a situation in which the stereotypes and the facts part company (1961 ed., p. 111)." These strongly stated postulates contain the kernel of moral indigna-tion that i s so characteristic of much of the later writings on stereotypes. In turn, the examples that Lippmann employed to i l l u s t r a t e his basic argument contain the seeds of each of the three major research orientations. 14 To a large extent they were based on war anecdotes and, because of this, they were often gross overgeneralizations which were less than rational in addition to being highly consensual. It has been d i f f i c u l t , because of this firmly established legacy, to give adequate consideration to those stereotypes which may be favourable (e.g., Hudson, 1968, 1970) and flexible (e.g., Dudycha, 1942; Seago, 1947), and which may contain a "kernel," i f not more, of truth (e.g., Abate and Berrien, 1967; Schuman, 1966). Indeed, in order to do so one must f i r s t attend to the arguments against the use of one or a l l of the "overgeneralization," " i r r a t i o n a l , " and "consensual belief" perspectives as the sole c r i t e r i a for defining stereotypes. Additionally, one must question the adequacy of a neo-Platonic perspective which assumes that non-specialists, and particularly non-academic non-specialists, are cognitively inept. In view of the widespread research emphasis on negative stereotypes i t i s ironic to note that Lippmann, himself, made reference to the value of a broad, general orientation toward the f i e l d . As we have seen, however, the constraints of his socio-political thesis prevented him from pursuing any issues that did not bear directly upon the undesirable facets of the stereotyping process. 15 Stereotypes as "social" concepts From a broad perspective which, encompasses complimentary and non-commital as well as derogatory characterizations, stereotyping can be seen to be a psychological process which i s at least as much functional as i t is dysfunctional. Stereotypes order the social environment of the individual into a coherent system or systems. They allow a person to verbalize and to conceptualize about his feelings toward the target of the stereotype, to mentally rehearse expected social episodes, to reanalyze previous encounters, and, in specific instances, to adopt what he considers to be a suitable behavioural orientation. Thus, the socially adjusted individual may employ stereotypes which conform closely to the (generally agreed upon) dimensions underlying social reality, while the more prejudiced person may, indeed, use stereotypes in much the same manner as Lippmann and his adherents envisaged. In essence, then, stereotypes should properly be regarded as concept [s] ..., with positive as well as negative functions, having the same general kinds of properties as other concepts, and serving to organize experience as do other concepts, (Vinacke, 1957, p. 229) Stereotypes can be distinguished from other concepts, not because they suffer from any moral or logical inadequacies, but rather because they incorporate personality t r a i t terms as well as terms pertaining to non-social, physical characteristics (Vinacke, 1957). This defines them as concepts which are specifically social i n nature. According to this definition, then, any group or individual i n the socio-cultural environment can become the target of a stereotype. In 16 addition, stereotypes can refer to role and personality descriptions that may be f i l l e d , at least potentially, by a person or a group of people. Thus one can hold a stereotype about Canadians, Southern Baptists, Charlie Chaplin, a good (or a bad) member of parliament, and a team of heart surgeons, for example. On the other hand, one's concept of England i s not a stereotype. The term encompasses not only the perceived characteristics of a national group, but also features of the terrain, classes of flora and fauna, aspects of the natural resources, and so forth. It is too inclusive a concept to be a stereotype. For similar reasons one cannot have a stereotype about fascism, a particular set of p o l i t i c a l ideologies, but one can have a stereotype about fascists, the adherents to this ideology. In sum, the proposed definition of stereotypes encompasses the f u l l range of individuals' concepts of the various people and groups which have existed, do exist, or can potentially exist in the social environment. It does not refer to conceptualizations of the values, institutions, and environmental settings which form a framework for, and a backdrop to, the social interactions of these people and groups. 17 Implications of the "social concept" definition of stereotypes There are a number of advantages to be gained from adopting this broad definition of stereotypes. Fundamentally, the definition makes explicit what has t a c i t l y been assumed since the early work of Rice (1926), Litterer (1933), and Katz and Braly (1933): Namely, that stereotypes refer specifically to performers within the social environment. It also raises the possibility that this set of concepts differs from others not only i n i t s potential to incorporate personality items into i t s descriptive lexicon, but also i n terms of a number of structural properties such as c l a r i t y , permeability, and the inclusion of uncommon and idiosyncratically used words. More importantly, however, i t releases researchers from the efforts involved i n defending and bolstering overly restricted perspectives which are more complementary than antithetical. By so doing, i t allows them to expend greater energies on distinguishing between different classes of stereotypes, and upon isolating the processes underlying, and the characteristics defining, these classes. For instance, one can explore the situational and personality correlates of those (favourable and unfavourable) stereotypes which are open-textured, flexible, and sensitive to changes in environmental conditions, as well as those which are closed and resistant to modification. Similarly, one can attempt to isolate the variables which determine whether or not stereotypes are formed from uncommon and idiosyncratic items, or from consensual and widely subscribed to terms. One can also 18 work toward marking those features, i f any, which distinguish between con-cepts of roles and character types and those of 'real' and familiar people, between concepts of groups and those of individuals, and between the concepts of targets that we "know about" (Kelvin, 1970) and the concepts of targets we 'know through personal experience.' In addition, when considering the issues pertaining to stereotype formation and to the conditions and manner in which stereotypes are used and acted upon, the general perspective offered by the present definition readily allows one to attend to related areas of research concerning impression formation, attribution theory, and the structure of cognitive dictionaries and semantic memory. Yet, whatever new avenues of research may be developed, one can safely predict that the traditional focus upon faulty and inaccurate stereotypes w i l l continue to command strong interest. After a l l , these characterizations readily stand out from others as potential threats to the harmony of social interactions. One can expect, too, that the standard practice of using groups rather than individuals as target stimuli w i l l be maintained. Indeed, i t would be surprising i f these research interests were not continued, since i t was the attention given to such longstanding issues which was responsible for the establishment of the stereotype f i e l d in the f i r s t place. 19 2. Stereotypes and Social Groups To this point in the discussion we have followed Lippmann's example and treated stereotypes specifically as psychological entities, as those conceptual profiles through which we make order out of our experiences and social encounters. Yet, although he did not acknowledge the fact, there were at least two levels on which Lippmann treated sterotypes as s t r i c t l y sociological phenomena. It i s worth pursuing these lines of thought further, since the failure to distinguish between the different  types of stereotypes has, in i t s turn, resulted in some conceptual confusion within the f i e l d . Cultural Stereotypes Researchers who accepted Lippmann's propositions found i t d i f f i c u l t to imagine that 'rigid,' 'oversimplified,' and 'prerational' stereotypes could be the products of quite normal cognitive processes. Rather, there was general agreement that stereotypes were borrowed, ready made, from the commonly available "pictures" that are embedded in the media, in folklore, and in other purveyors of popular culture (e.g., Hayakawa, 1950). Consequently, those writers who failed to concede that there was more than one "genus" of stereotype equated the stereotypes that can be found in sources such as newspapers, lit e r a r y works, and p o l i t i c a l propaganda with those that are a part and parcel of people's cognitive systems. (Eysenck and Crown (1948) were notable exceptions to this rule.) The stereotypes that are contained i n the media and the arts are free-floating in the sense that they do not depend upon any one individual 20 or sub-cultural group for their existence. Although they have been created by various high-status opinion generators (e.g., journalists, politicians, social c r i t i c s , and t r i b a l elders) they have become dis-sociated from these people i n the process of being projected into the prevailing 'world view.' In being assimilated into the social heritage these stereotypes take on a distinctly cultural form. That i s , they can  be considered and discussed without reference to any particular person's system of beliefs and values. The news media 'portraits' of groups such as the I.R.A. and the F.L.O. are particular examples. People may borrow a l l or part of the content of these cultural  stereotypes i n forming their own particular concepts of a target. Nonetheless, whatever the similarity in content, cultural stereotypes, as social entities, are logically different from the personal stereotypes (Secord and Backman, 1964), or mental entities, which may be modelled on them. Dwelling further on this distinction, i t becomes clear that there are many possible instances i n which a personal stereotype may have nothing i n common with i t s cultural 'counterpart' (or counterparts). As a matter of fact since only a limited set of individuals and groups ever achieve social recognition, there are numerous cases where personal stereo-types are formed in the absence of any cultural "picture." We may have a well-defined concept of our high-school, music-freak friends, but i t is unlikely, unless of course they became famous, that we could find a description of the group in any of the repositories of cultural information. 21 Generally speaking, then, i t i s only when a target has achieved some measure of social recognition that we can expect to find one or more cultural counterparts to an individual's personal stereotype. Furthermore, we can expect to find marked similarity between cultural and personal stereotypes only when the individual's experiences with the targets correspond closely with the experiences of the opinion generators respons-ible for producing the cultural stereotypes. Whether we agree or disagree with them, however, the cultural stereotypes to which we most readily attend are those which receive a wide measure of support. They are the ones that relate to the experiences of a broad section of the population. Because of this, they are highly visible i n the media and the arts. When they f a i l to be of relevance to the community at large they disappear. They either cease to exist, as is the case in pre-industrial societies founded on an oral tradition, or, as i s the case i n technological societies, they become lost i n manuscript and archival collections to be sought out only by historians and other like-minded academics. Social Stereotypes When we speak about cultural stereotypes receiving a wide measure of support we return, once again, to the question of consensus. It i s often the case that when a particular target receives a great deal of public attention reference groups play an important part in the way people structure and interpret the available information (Diab, 1962). For example, immediately before Richard Nixon resigned from the Presidency 22 of the United States the American population was sharply divided between those who supported the pro-impeachment lobby and those who supported the anti-impeachment group. Different cultural stereotypes of Richard Nixon (and the presidential role) emerged from each of the two "camps." Depending upon their position i n the debate, people referred to and borrowed from one or the other of the contrasting sets of cultural stereotypes in the process of forming, or articulating, their own personal stereotypes of Richard Nixon. Indeed, there was a great deal of agreement between the anti-Nixonites with respect to their personal stereotypes of Nixon, just as there was between the pro-Nixonites. If an investigator had used the traditional methodology of stereotype research, had selected a sample of 'pro-impeachment' people, and had asked each one in turn to characterize Richard Nixon he would have found that there were a number of words which frequently reoccurred in the various descriptions of the target and which seemed to form a 'summary picture' of the group's beliefs about Nixon. The summary picture, i n outlining the general group perspective, would have defined the "social norm" for describing the target (cf. Karlins, Coffman, and Walters, 1969, p. 3). Because of the degree of cohesion among the group, this social stereotype (Secord and Backman, 1964) would have borne a marked similarity to the relevant cultural stereotype(s) and to the personal stereotypes from which i t was abstracted. Thus, the summary pictures provided by social stereotypes can be quite useful when discussing groups' perspectives as well as their orientations toward each other. 23 It must be remembered, however, that social stereotypes do not take account of the uncommon and idiosyncratic elements of personal stereotypes. Indeed, the attributes which any one individual considers to be most characteristic of a particular target may not be represented at a l l i n the cluster of most popular responses. Furthermore, even when there i s some relationship between the components of a personal and social stereotype the degree of overlap may not be extensive. It becomes clear, therefore, that there can be marked differences among social stereotypes with respect to how well they encompass and represent the personal perspectives of a l l of the members of the respective sample groups. For instance, a group may focus sharply on one target but not pay a great deal of attention to another. The group members would be exposed to a highly visible set of cultural stereotypes and would share i n common a great many experiences pertaining to the f i r s t target, and, as a result, the social stereotype would likely be sharply defined and would reflect the general agreement among the individuals within the group. However, the group members would have neither a common orientation toward the second target nor a salient cluster of cultural stereotypes from which to borrow information. Their personal stereotypes would grow out of their often limited and very different experiences with the group or individual in question. Consequently the social stereotype, reflecting this lack of communal focus, would be vague and weakly defined. The structure of social stereotypes can be affected not only by the intensity of the group focus on the target but also by the type of group 24 from which they are derived. The individuals i n two different groups, for example, may hold equally strong personal stereotypes about a target but one social stereotype may be.much more ambiguous and less consensual than the other as a function of the group composition. For example, the members of one group may be a great deal more heterogeneous than those of the other with respect to their personal characterizations of the target. Thus, the social stereotype of Nixon derived from a group of Nixon's Orange County supporters would be much more clearly defined than one derived from a random sample of Americans who could be counted upon to subscribe to a wide variety of p o l i t i c a l and social views. In short, some social stereotypes are more li k e l y than others to mirror their personal (and cultural) counterparts. As a consequence, one must be careful not to treat every one of them as i f they were equally precise, equally representative, and equally useful summary pictures of the personal stereotypes of group members. This i s particularly so in the case of samples drawn from such arbitrary populations as introductory psychology students. Taking account of these cautionary arguments, however, we can attend to the favourableness of the list e d characteristics and to the frequency with which they were employed, and, i n a less exact sense, use social stereotypes as general indexes of the consensus between group members concerning their characterizations of, and attitudes toward the specific targets. When social stereotypes are used i n this sense one can consider the traits l i s t e d to be examples of the types of terms that one could expect 25 to find i n the personal stereotypes of many of the individual group members. Social stereotypes, then, are unlike their personal and cultural counterparts i n that the latter are 'real' social and psychological entities while the former are research-created metaphors. Cultural and social stereotypes as a context When we discuss cultural and social stereotypes we not only confront some definitional issues which have caused problems i n the past but we also provide a broad sociological context for the study of personal stereotypes. Indeed, interest i n personal stereotypes l i e s just as much in explor-ing their environmental determinants as i t does i n elaborating upon their structural properties. To this end i t i s important to determine the ways in which personal stereotypes are influenced by group norms and other social variables, and the extent to which they reflect the prevailing cultural climate and conform to the related concepts of individuals within the same social groups. The analysis of cultural and social stereotypes can help In the achievement of these goals. For example, by mapping the cultural stereotypes of a particular national group one can determine which social targets, both within the group and without, receive a great deal of attention at any particular time. One can also obtain a historical perspective on the changes i n national outlook and the changes in cultural stereotypes that have occurred over specific time periods. Curtis (1971) has done this admirably in the case of the changing Victorian cultural stereotypes of the Irish. 26 In addition, by looking at the way cultural stereotypes cluster, complement each other, and are antithetical to each other one can get a sense of the values and opinions of the various sub-cultural groups which enter into the debate over the representation of specific targets. The Vietnam War era in the United States, for instance, provides many examples of conflicting cultural stereotypes which grew out of considerable social discord. Social stereotypes can be used to add a further dimension to this framework of representations. As the previous discussion has indicated, when carefully constructed they can reflect the amount of agreement between, and the general orientation of the members of well-defined, sub-cultural groups in characterizing specific targets. Broadening the perspective of stereotype research to allow a simultaneous consideration of personal, social, and cultural stereotypes w i l l have a number of desirable consequences. It w i l l require the development of comprehensive, interdisciplinary models that can encompass research findings from a wide variety of sources. It w i l l provide a necessary social context for research and in so doing w i l l permit the testing of hypotheses which could not adequately be addressed i n a laboratory environment. Finally, because of the emphasis on context, i t w i l l help to determine whether the processes involved in stereotyping are specific to local settings, as Thorngate (1975) suggests most social psychological processes might be, or whether there are invariances in the stereotyping process which reliably reoccur in different social settings in different historical periods. 27 Summary Traditionally, stereotype studies have fallen squarely within the framework of prejudice research. In so doing they have tended to cluster around three major theoretical orientations which depict stereotypes as being either overgeneralizations, consensual beliefs, or r i g i d and irrational substitutes for thought. Although these view points can be c r i t i c i z e d for having too limited a perspective they continue to maintain their popularity and to be widely subscribed to. The general orientation of the stereotype f i e l d has grown out of the writings of Walter Lippmann (1922). He argued, within the context of a broad p o l i t i c a l thesis, that stereotypes are fixed and prerational images in our heads which can serve to distort our perceptions and cause us to make erroneous interpretations of our social world. His i l l u s t r a t i v e examples contained the seeds of each of the "overgeneralization," "prerational," and "consensual belief" perspectives which have dominated the f i e l d subsequently. Thus, i t is only be taking account of the limitations of Lippmann's p o l i t i c a l points, by challenging his neo-Platonism, and by attending to the arguments directed against each of the three major theoretical orientations, that one can begin to get a sense of the more inclusive framework which can serve to encompass and lend new directions to stereo-type research. From a broad perspective stereotypes can be defined as social concepts which differ from other concepts i n that they can include personality terms as well as terms referring to non-social, physical characteristics 28 (Vinacke, 1957). In essence, then, they are concepts about groups and individuals, the performers In the social environment. When discussing stereotypes one must be careful to distinguish between personal stereotypes, which, as concepts in our heads, are psychological entities, and cultural and social stereotypes. Cultural stereotypes are those characterizations which are embedded in the media and other sources of cultural information, while social stereotypes are research-generated indexes of the consensus between sample group members concerning their characterizations of, and general orientation toward specific targets. If used appropriately, cultural and social stereotypes can provide a rich social context for the study of personal stereotypes. The inter-disciplinary perspective that such a research approach would require may produce some f r u i t f u l insights into the general process of stereotyping. * * * 29 CHAPTER TWO: MEASURING INSTRUMENTS AND  STEREOTYPE RESEARCH2 The adjectival check l i s t : Some problems A variety of structural (e.g., Brigham, 1973; Gardner, Kirby, Gorospe, and Villamin, 1973; Tanaka, 1965), semi-structured (e.g., Bjerstedt, 1960), and open-ended (e.g., Cahalan and Trager, 1949; Lambert and Klineberg, 1959; Reigrotski and Anderson, 1959) measuring instruments have been introduced into the f i e l d of stereotype research in order to determine which characteristics people employ when describing particular social targets. The adjectival check l i s t (e.g., Katz and Braly, 1933) has always been the most widely used of these techniques. Thus, respon-dents i n stereotype studies are typically asked to choose from a l i s t of seventy-five or more traits those terms which they think best characterize the group or individual i n question. Despite i t s popularity as a research tool the adjectival check l i s t procedure has not escaped c r i t i c a l attention. Indeed, subjects themselves were the f i r s t to raise objections, and Bjerstedt (1960), Eysenck and Crown (1948), and Gilbert (1951) are among those who have reported that many of their respondents voiced reservations about, or even refused to complete the check l i s t tasks on the grounds that they were irrelevent and meaningless. Taking note of these criticisms, some researchers have surmised that the words in any particular check l i s t may not be ones that the subject i s familiar or comfortable with (e.g., Duijker and Frijda,1960). 30 Others have pointed out that i t is often very easy for subjects to provide what they know to be socially desirable responses (e.g., Gilbert, 1951; Sigall and Page, 1971), to respond even when they have no particular views on the subject (Brown, 1965; Eysenck and Crown, 1948), and, perhaps, to provide edited versions of media portraits and other cultural stereo- types which do not conform to their own personal view points (e.g., Eysenck and Crown, 1948). In short, there i s an undercurrent of opinion which suggests that the characterizations produced from check l i s t s (and other structured stereotype-eliciting tasks) may not be good approximations of those used by individuals in the course of their normal everyday l i v e s . The appropriateness of stereotype measures It may be presumed that i n the process of stereotyping, a person carries out a memory scan for appropriate words and phrases with which to describe the target. From this short l i s t of salient items he chooses those which he finds most relevant to a description of the personality and social characteristics of the group or individual under consideration. This i s the case whether the terms are favourable or unfavourable; and whether they are idiosyncratic, resulting from his unique experiences with the target i n question, or borrowed from the cultural definitions of which he i s at least marginally aware. The individual's particular characteri-zation may vary somewhat from social setting to social setting, but, i n any case, the l i s t of items that he provides i s recalled from the pool of tr a i t terms in his cognitive dictionary. 31 Consequently, check l i s t s can produce stereotypes which mirror those normally used by people only in so far as they contain t r a i t items that are salient to the respondents and that are commonly used by them. The extent to which this degree of relevance is built into any particular l i s t i s a matter for empirical determination. Yet, on an intuitive basis, one must suspect that the items in check l i s t s do not overlap to any great extent with the terms within individuals' cognitive dictionaries. It i s easy, for example, to imagine a situation in which respondents are forced to substitute a word as an approximate alternative to a preferred term which does not appear i n the provided l i s t of seventy or more items. S t i l l more probable i s the implicit reaction: "I didn't think of that word but, by golly, i t sure does describe the group. I must check i t . " In both cases such traits are l i s t e d on the basis of recognition rather than r e c a l l . The more they predominate among individuals' responses the greater w i l l be the qualitative differences between check l i s t stereotypes and those produced in naturalistic settings. Because they can be based so heavily on recognition, check l i s t characterizations do not seem to be the appropriate analogues of spontaneously produced stereotypes. Rather, they appear to be more similar to those representations which we readily endorse as extensions of our own personal stereotypes after they have been brought to our atten-tion in the course of conversations, lectures, and other information generating a c t i v i t i e s . Indeed, stereotypes which are freely produced i n natural social settings are better measured by techniques which require the re c a l l rather 32 than the recognition of descriptive terms. Although they are seldom used, the open-ended procedures, such as unstructured interviews and free response paper and pencil tests, would seem to f i l l this requirement. Relevant research findings If structured and unstructured stereotype tasks do indeed reflect different psychological processes they should produce markedly different research findings. There is evidence to indicate that this i s the case. Ehrlich and Rinehart (1965) found that American undergraduates who had been asked to characterize six national groups with a check l i s t employed a greater number of words than respondents who had been asked to use an open-ended format. They found also that the check l i s t respondents achieved a much higher level of interpersonal agreement concerning the assignment of specific characteristics to particular targets than did the free response subjects. Furthermore, the open-ended task yielded a greater variety of tr a i t terms with which to describe the national groups than did the structured task, and i n a l l , only 13% of the reported t r a i t descriptors emerged in common from both measuring formats. These findings were corroborated by McTiernan (1973) in a study that employed a repeated measures rather than a between groups design. He asked a sample of Irish students to characterize eight national and subnational groups from the British Isles using both measuring instruments. Once again, there was a much greater degree of consensus between respon-dents i n choosing check l i s t characteristics with which to portray the groups than in describing them in their own words. Once again, too, each 33 of the eight pairs of check l i s t and free response social stereotypes (Secord and Backman, 1964) offered complementary rather than common portraits of the targets. In addition, the personal stereotypes derived from the structured instrument were not only considerably more differen-tiated than those obtained from the open-ended format, they were also, generally speaking, more favourable. However, not only did the data from the structured and unstructured measuring techniques convey very different impressions concerning the differentiation, favourableness, and content of the personal stereotypes, they also allowed for very different decisions to be made concerning the similarities and differences between the characterizations of the various targets. For instance, the check l i s t data indicated that the personal stereotypes of two of the four subnational targets were more highly differentiated than those of the remaining two groups. Yet, the free response data yielded no such finding. Similarly, the check l i s t personal stereotypes of the four national groups did not diffe r with respect to their favourableness, although some of these targets were portrayed less favourably than others on the open-ended format. Core and peripheral stereotype components and their measurement The findings of the McTiernan (1973) and Ehrlich and Rinehart (1965) studies form a particularly solid body of evidence to support the hypo-thesis that different processes predominate in people's responses to check l i s t and open-ended stereotype tasks. 34 Elaborating upon the contention that the free response characteriza-tions are based on recall and the check l i s t stereotypes are founded, to a great extent, on the recognition of pertinent t r a i t s , one can argue that the open-ended format taps the core components of individuals' personal stereotype systems while check l i s t s provide a sample of the peripheral  elements. As previously indicated, however, check l i s t responses can contain core elements of personal stereotypes in addition to peripheral components. Thus, in their traditional form, check l i s t s do not provide an uncontam-inated representation of those traits which are on the periphery of a person's concept about a target. For them to more clearly reflect the peripheral elements of personal stereotypes i t i s necessary, as a matter of procedure, to compare check l i s t responses with those previously obtained on an open-ended format and to eliminate a l l of the checked items which: already appear among the free response characterizations. Implications of a revised measuring procedure The implications of the conceptual and empirical distinction between the core and peripheral components of personal stereotypes seem clear. The traditional check l i s t research has not provided a complete account of the content and structure of personal and social stereotypes. By f a i l i n g , i n large measure, to account for the core components of stereotypes i t has presented us with a considerable mound of data which is more intuitively appealing than psychologically valuable. Indeed, as Brigham (1971) and others have indicated, i t has been d i f f i c u l t to 35 delineate the relationships between personal stereotypes and general attitudes and behaviour. It may well be that these d i f f i c u l t i e s have arisen, not because the relationships do not exist, but rather because the measuring instruments did not focus on the central and most relevant components of stereotypes. By attending more closely to the core components one can provide a more balanced picture of the content of personal and social stereotypes, and, at the same time, one may more readily be able to integrate and to relate stereotype research to the findings i n other areas of social and cognitive psychology: A policy advocated by Taylor and Aboud (1973) and currently being implemented by Aboud (1976). The emphasis on open-ended research formats that i s necessitated by such an orientation, and their use In conjunction with appropriate structured instruments, is not without methodological problems. The homogeneity of variance assumption can easily be violated, for example, and this can offer some s t a t i s t i c a l d i f f i c u l t i e s when working with unequally sized samples (cf. Petrinovich and Hardyck, 1969). However, the practical as well as the theoretical advantages of this extended measuring format compensate for these complications. Furthermore, when used on their own, free response techniques by their nature are highly portable instruments. They can readily be used in situations other than classrooms and social psychology laboratories. Sun (1976) has found, for example, that unstructured stereotypes e l i c i t e d over the phone are comparable to free response characterizations collected under more traditional conditions. By extending the use of such formats 36 to factories and streets, to picket lines and community meetings, one can go a long way toward reintroducing the notion of context to stereotype research. It i s , after a l l , i n just such situations that stereotypes may be expected to mediate in the processes of attribution, social perception, and interpersonal behaviour. Summary Despite the introduction of numerous measuring instruments to the f i e l d , the adjectival check l i s t continues to be the most widely used technique i n stereotype research. However, a number of criticisms directed against the procedure have raised the possibility that check l i s t characterizations may not be appropriate analogues of the stereotypes used by individuals in the course of their everyday live s . The issue revolves around the argument that freely produced stereo-types are based on r e c a l l , while check l i s t stereotypes are based on recognition and, as such, are the products of a different psychological process. From this assumption one can further contend that freely produced stereotypes are better measured by unstructured, open-ended instruments, while check l i s t characterizations can usefully serve as indexes of those representations which we are prepared to endorse as extensions of our own personal stereotypes after they have been brought to our attention in the course of conversations and other social a c t i v i t i e s . Indeed, research evidence indicates that free response and check l i s t procedures convey very different impressions concerning the favourableness, differentiation, and content of personal stereotypes. They also allow for 37 very different decisions to be made concerning the similarities and d i f f e r -ences between the characterizations of the various targets. Such findings add support to the assumption that responses to structured and unstructured stereotype tasks are governed by different psychological processes. Since they are based on recall the open-ended measuring procedures can be said to tap the core components of personal stereotypes. On the other hand, check l i s t s , which are founded on recognition, provide a sample of the peripheral elements of stereotype characterizations. Thus, in order to obtain a more complete understanding of the nature and contents of personal stereotypes i t i s necessary to use a combination of an open-ended and a modified check l i s t procedure. This form of instrument has many advantages, not least being i t s adaptability for easy use outside of classroom and laboratory settings. * * * 38 P A R T TWO KNOWLEDGE OF THE TARGET'S CHARACTERIZATION OF THE RESPONDENTS' NATIONAL GROUP AND THE STEREOTYPE FORMATION PROCESS 39 CHAPTER THREE: A PRELIMINARY HYPOTHESIS CONCERNING  RECIPROCITY AND STEREOTYPE FORMATION Historically, stereotypes have been treated in static rather than dynamic terms. Researchers have been interested primarily i n defining their contents and l i t t l e systematic work has been conducted to determine the variables that govern their growth and development. Stereotype formation and the social Context Workers who have addressed the question of stereotype formation have generally followed Lippmann's (1922) example and have focussed upon the influence of the social environment. It has been argued, for instance, that information about social targets i s obtained largely from hearsay and gossip (Klineberg, 1964), from the mass media (Tanaka, 1972; O'Hara, 1961), and from the school environment (Fishman, 1956; S t i l l w e l l and Spencer, 1973). Each of these assertions i s founded on the more general observation that cultural and subcultural reference groups project well-established "social norms for describing recognized groups of persons" (Karlins, Coffman, and Walters, 1969, p. 3). Indeed, such norms can exert consider-able influence on individuals when they stereotype particular targets (e.g., Avigdor, 1953; Diab, 1962; Sherif, Harvey, White, Hood, and Sherif, 1961) and their influence i s especially evident i n the responses provided by children who have been asked to describe various ethnic and national groups. 40 It has been demonstrated repeatedly that children as young as four or five are aware of group differences (Clark and Clark, 1947; Goodman, 1952; Radke, Trager, and Davis, 1949), that by six or seven they are capable of characterizing target groups with t r a i t terms, although not with any great degree of consensus (Lambert and Klineberg, 1959), and that as they grow into adolescence the amount of interpersonal agreement increases and their stereotypes become more and more similar to those of the adult community (Blake and Dennis, 1943; Gardner, Taylor, and Feenstra, 1970; Kirby and Gardner, 1973). Psychological processes and the development of stereotypes These developmental changes in the content and structure of personal stereotypes can be accounted for by at least two processes, each of which involves social learning. As children become familiar with the general rules of social inter-action they learn which set of t r a i t terms are commonly associated with which particular targets (e.g., Blake and Dennis, 1943; Kirby and Gardner, 1973; Pettigrew, 1966). However, associative learning of this sort is restricted to those instances where stereotyped targets possess clear cultural definitions. Consequently individuals must also acquire general, open-ended "stereotype construction" strategies which are capable of accommodating a broad range of novel as well as familiar stimuli. The acquisition of a differentiated t r a i t lexicon is essential to the development of these strategies. In addition one must learn the 41 cultural rules for associating t r a i t s , for classifying targets, and for attributing particular t r a i t clusters to specific targets on the basis of the classificatory c r i t e r i a . It can be assumed that these cognitive operations are learned during the period in which individuals acquire the basic s k i l l s of social inter-action and interpersonal behaviour. A considerable body of information has already been obtained about the nature of the various classificatory and associational rules (cf. Cook, 1971). For example, the manner in which individuals emphasize, combine, and organize traits within their "implicit theories of personality" has been explored extensively (e.g., Asch, 1946; Wishner, 1960; Bruner, Shapiro, and Tagiuri, 1958; Koltuv, 1962; Lay and Jackson, 1969; Norman, 1966). Research has also identified many of the social and quasi-social markers, from physique, the use of make-up, and gaze to sex, nationality, culture, and religion, that are frequently used to identify and distinguish between individuals and groups ?(e.g., Strongman and Hart, 1968; McKeachie, 1952; Kendon and Cook, 1969; E l l i s and Bentler, 1973; Katz and Braly, 1933; Jones and Ashmore, 1973; McTiernan, 1973). Furthermore, stereotype studies have indicated the types of t r a i t clusters that are associated with targets identified solely on the basis of information such as age, sex, nationality, race, and socioeconomic status (e.g., Sun, 1976; Broverman, Broverman, Clarkson, Rosenkrantz, and Vogel, 1970; Kaflins, Coffman, and Walters, 1969; Brigham, 1971; Feldman, 1972). They have determined also that when targets are defined in more complex terms some of the category labels (for example, 42 the occupation, social mobility, and role items) carry more weight than others (for example, the race and geographic variables) in shaping the form and structure of the e l i c i t e d stereotypes (e.g., Aboud and Taylor, 1971; Aboud, Taylor, and Doumani, 1973; La Gaipa, 1971; Feldman, 1972; Feldman and Hilterman, 1975; Taylor, B a s s i l i , and Aboud, 1973). A preliminary model of the stereotype formation process This broad set of research findings provides us with a foundation for speculating upon specific strategies that individuals may employ when forming stereotypes. The evidence concerning the use of category labels suggests that our i n i t i a l response to a social target may be to build as complete and specific a demographic profile of i t as possible. We achieve a degree of specificity with respect to complex stimuli by weighing some pieces of information about the targets more heavily than others. Our weighting c r i t e r i a are determined in part by the socio-cultural context in which the targets are being evaluated (e.g., Aboud, Taylor, and Doumani, 1973). They are reflected also in our personal rankings of the relative importance of the various category labels (cf. Feldman and Hilterman, 1975). In those instances in which the target i s defined in general and simplistic terms we elaborate upon our profile by attributing additional category labels to the stimulus group or stimulus person in question (e.g., Bayton, McAlister, and Homer, 1956; Kelvin, 1970). La Gaipa (1971), for example, demonstrated that respondents readily associate particular 43 occupations with, different ethnic groups, and they do this, very often, with a substantial amount of interpersonal agreement. Having classified targets demographically we then proceed to attribute t r a i t clusters and personality terms to them. Kelvin (1970) has noted that many of these descriptive terms refer to particular attitudes (e.g., tradition loving, superstitious) as well as to behaviours (e.g., industrious, talkative), and general dispositions (e.g., intelligent, impulsive). The use of such attitude terms, especially the more specific items like "loyal to family ties" and "extremely nationalistic", indicates that the respondents are aware of, or at least can estimate in a general sense, many of the beliefs, values, and perspec-tives of the target groups or target persons. On a p r i o r i grounds i t seems most probable that inferences about, or an awareness of targets' social orientations and world views occur subse-quent to their being classified demographically and prior to their being stereotypes with t r a i t terms. This suggests that individuals' impressions of a target's general views of the world may be important mediating variables i n the stereotype formation process. Indeed Rokeach, Smith, and Evans (1960) have argued, for example, that black Americans have tradi-tionally been the victims of prejudice and derogatory characterizations not because of their race per se but rather because their beliefs were assumed to be dissimilar to those of whites. Incorporating the arguments of the previous paragraphs into a three-step model we can hypothesize that as a f i r s t step toward forming their own personal stereotypes individuals construct demographic profiles of 44 targets. They then associate, particular world views and value systems with these profiles, and subsequently generate what they consider to be appropriate sets of tralt-attitude-behaviour attributions on the basis of these inferences about the targets' social perspectives. It can be assumed further that individuals w i l l employ the outlooks and orientations of their own salient reference groups as judgemental anchors when evaluating the perspectives and attitudes of the various targets to which they attend. The res earch p roblem In order to test this model It is important to determine the types of value and general world view dimensions that may prove to be highly salient i n the stereotype formation process. It i s then necessary to establish: (1) whether or not individuals make inferences about targets' positions along these dimensions in the absence of relevant information, (2) whether or not information about targets' positions along these dimensions is attended to when embedded within the context of more general profiles of the targets in question, and (3) whether or not such information influences the form and structure of the emergent personal stereotypes. Some of the dimensions that may be focussed upon by individuals when making inferences about targets' social orientations have already been defined by research. Predictably these Include such salient variables as religious belief and p o l i t i c a l perspective (e.g., Rokeach, Smith, and Evans, 1960; McTiernan, 1973; Jones and Ashmore, 1973). 45 In addition, person perception studies have isolated another variable which may play a significant role in the stereotype formation process. Work on interactions in dyads and groups has indicated that respondents' reactions to target persons were greatly influenced by their perceptions of how these target persons viewed them in return (e.g., Tagiuri, Blake, and Bruner, 1953; Tagiuri, 1958; Secord and Backman, 1964). Moreover, individuals can readily,.and in some cases accurately estimate the types of stereotypes held by the members of different ethnic, age, and national target groups about their own particular ethnic, age, and national reference groups (e.g., Kaplan and Goldman, 1973; Sun, 1976; McTiernan, 1973). Thus the relationship between individuals' "perceptions of how others feel toward them and their own feelings for others" (Tagiuri, Blake, and Bruner, 1953, p. 585) may generalize to instances in which groups rather than individuals are being characterized; and the content and structure of respondents' stereotypes of target groups may be deter-mined in part by their perceptions of the targets' characterizations of their own salient reference group (or groups). The present study was designed to address this latter proposition and to explore the functional relationship between the development of respondents' stereotypes toward an unfamiliar target group and their perceptions of how this target group characterizes their own national reference group. 46 By exposing individuals to one of three versions of a documentary presentation about a previously unknown group, the Orkney Islanders, i t was possible to answer the following questions: Will respondents make inferences about the Orkney Islanders* stereotypes of their own national group in the absence of definite information about the nature of these "reciprocal stereotypes"? Will such information be attended to when i t is included in a documentary script about the Islanders, and w i l l i t affect the form and structure of the core and peripheral components (see Chapter 2) of the newly developed stereotypes? In sum, w i l l the relationship between the respondents' personal stereotypes of the Orkney Islanders and their estimates of and information about the Islanders' reciprocal stereotypes of their own national group be such that we can support the general proposition that: An individual's inference about a target's characterization of his most salient reference group acts as a mediating variable in the stereotype formation process? In keeping with the findings and recommendations of previous stereo-type and person perception research, the core and peripheral components of the experimentally created stereotypes were examined in terms of their favourableness, their degree of differentiation, and their degree of commonness (cf. Brigham, 1971; Tagiuri, 1958). Social stereotypes were formed from the words which occurred most frequently i n the respondents' characterizations (cf. Karlins, Coffman, and Walters, 1969). These provided examples of the types of words that were most often included in the personal stereotypes. 47 Additionally, since the stimulus-materials included a detailed descrip-tion of the target group i t was considered potentially instructive to determine the degree of overlap between the stimulus material character-izations and the respondents' personal stereotypes. Summary Experimental findings as well as age-related developmental evidence suggest that the contents and structure of individuals' stereotypes are determined i n great measure by reference group norms. For instance, i n those cases in which there are clear group definitions of targets, individuals acquire their stereotypes through a straightforward process of social learning. Additionally, individuals depend largely on group-defined, t r a i t association and target identification rules i n order to form "stereotype construction" strategies that w i l l permit them to characterize novel and unfamiliar stimulus persons and stimulus groups. L i t t l e i s known about the nature of these stereotyping strategies. However, an examination of relevant research yields a preliminary three-step model of the general stereotype formation process. The model assumes that: individuals construct demographic profiles of targets, associate these with particular world views and value systems, and generate what they consider to be appropriate sets of t r a i t attributes on the basis of these inferences about the targets' social orientation. One further assumption i s that individuals use their salient reference groups as judgemental anchors when forming stereotypes i n this fashion. 48 Person perception studies indicate that individuals' perceptions of how a target characterizes their own reference group may be one such "world view" variable which mediates in the stereotype formation process. The present study was designed to explore this hypothesis by providing answers to the following questions: Will respondents make inferences about an unfamiliar target group's (the Orkney Islanders) stereotypes of their own national group in the absence of definite information about the nature of these "reciprocal stereotypes"? Will such information be attended to when i t i s included in a documentary script about the Orkney Islanders, and w i l l i t affect the form and structure of the core and peripheral components of the newly developed stereotypes? On the basis of previous research findings i t was decided to examine the experimentally formed stereotypes in terms of their favourableness, their differentiation, their degree of commonness, and their inclusion of descriptive terms contained in the stimulus materials. 49 CHAPTER FOUR; METHOD Overview Three experimental conditions were established by pairing a film about l i f e on the Orkney Islands with each of three versions of a tape recorded, documentary passage. The basic passage contained an account of the history, economic a c t i v i t i e s , and l i f e s t y l e of the people, in addition to a description of the Islanders which was built around twenty preselected, neutral t r a i t words (non-reciprocal condition). The other two variations of the documentary script included information concerning the manner i n which the Orkney Islanders "characterize Canadians". This information replaced nonessential "padding" material from the original script. The Islanders were described as viewing Canadians very favourably in the positive reciprocal stereotype condition, and twenty positive attributes were used as examples of traits which were frequently assigned to Canadians. Alternatively, i n the negative reciprocal stereotype condition the Orkney Islanders were described as viewing Canadians very unfavourably. The negative antonyms of the positive traits were used in this condition. The film-tape sequences lasted for twenty-four minutes each - the length of a short TV programme. Each version of the documentary was shown to a different group of respondents. The sequences were presented as rough copies of a film which was being prepared by the social science departments at the University of British Columbia for use in their adult education courses. The task of 50 each group of respondents, as i t was explained to them, was to evaluate and suggest improvements on a preliminary version of the film so that the f i n a l edition could be structured to f i t the students' and general viewers' interests as well as the teachers' needs. The individuals were asked to watch the film and list e n to the taped soundtrack carefully, and-then to assess the merits and shortcomings of the presentation on a questionnaire, which was distributed at the end of the viewing period. The questionnaire included items relating to the respondents' f i r s t impressions of the target group. There was a one week interval between the presentation of the stimulus materials and the measurement of the newly formed stereotypes. This time gap was planned in order to f a c i l i t a t e the organization and consolidation of the relevant information within the respondents' personal stereotype systems. The specific length of the interval was determined by the practical considerations involved i n convening and reconvening the experimental participants. In order to provide a general context for the stereotyping task the respondents were led to think that the second testing session was part of an extensive research project concerned with obtaining norms on the manner In which Canadians characterize a wide variety of social groups, and on the way in which Canadians think other groups view them. The emphasis on the Orkney Islanders as a target was explained as being only incidentally related to the previous week's film session. Thus, unt i l the f i n a l debriefing, every effort was made to ensure that the respondents believed that they had participated in two distinct experiments. 51 The participants in the study were volunteer cadets from the National Sea Cadet Camp at H.M.C.S. Quadra, Comox, B.C. One hundred and ninety-three respondents completed a l l stages of the experiment. Fifty-eight saw the positive reciprocal film, sixty-seven saw the ribri-recipfocal documentary, and sixty-eight were shown the negative reciprocal film. One week after they had been exposed to the stimulus materials and had completed the post-film questionnaire the respondents were recalled to the " f i e l d laboratory" and asked to complete a four-part stereotyping task using a free response as well as a check l i s t instrument. The participants were required i n i t i a l l y to characterize Canadians, Orkney Islanders, Germans, and Arabs in their own words and then to repeat the task while playing the role of an Orkney Islander. In addition they were asked to rate each of the terms that they had listed on a nine-point favourableness scale. Finally, the entire free response sequence was repeated with the adjectival check l i s t format. The order of presentation of the targets was randomized within each of the four sections of the task. The check l i s t contained the twenty positive, neutral, and negative t r a i t words that were included i n the stimulus materials, as well as fifteen other words that were selected from the Karlins, Coffman, and Walters (1969) adjective l i s t . 52 Stimulus materials 1. The film-tape sequences The Orkney Islanders were chosen as a suitable experimental target for a number of reasons. The group was unfamiliar enough that i t could safely be assumed that few of the respondents, If any, knew anything about i t . At the same time i t was relatively easy to build plausible documentary scripts from information contained in a number of accessible sources. Indeed since the islands and the inhabitants played a historic role in the early economic and social development of Canada i t was possible to establish links between the Orcadians and Canadians without introducing a note of a r t i f i c i a l i t y into the stimulus materials. Finally, but of the utmost importance, suitable film footage about l i f e on the islands could be located. a. The film A twenty-four minute film, Bank Ahead (1967) , portraying the l i f e s t y l e and habitat of the Orkney Islanders was borrowed from the university's Instructional Media Centre and used as the visual focus of the study. The film traced the daily activities of a bank cashier and a seaman who worked together on a bank boat which serviced the financial needs of the outlying island communities. It dwelt at length on the interactions between the banker and the islanders and i t reflected the social as well as the economic functions of the floating banking service. More general passages were juxtaposed with, the bank boat scenes. These provided 53 background details about the islands,' terrain, the agricultural foundation of the Islanders' l i f e s t y l e , the various small export-oriented industries (e.g., whiskey d i s t i l l i n g and salmon smoking)., and the business and social l i f e i n Kirkwall, the major town on the main island. b. Primary script The basic script was constructed from information obtained from Encyclopaedia Britannica (1968), Collier's Encyclopaedia (1972), Bailey (1971), and the original soundtrack of the stimulus film, Bank Ahead (1967). The geographic location, population distribution, and principal features of the islands were described. Considerable emphasis was given to the "blend of modern and traditional values that govern the Islanders' optimistic outlook on l i f e and that are carried with them in their farming and business enterprises" (Appendix A, 1). The histo r i c a l l y changing relationship between the island communities and the outside world was discussed, and particular mention was made of the fact that many Orcadians emigrated to Canada and worked for the Hudson's Bay Company. It was pointed out in this respect that Canadians constitute one of many important and salient groups i n the world-view of the islanders. The Orkney Islanders, themselves, were characterized by a set of twenty evaluatively neutral t r a i t words (see Table 1). Each of the terms occurred twice in the text, once in adjectival form and once as an adverb. The order of presentation of the words was randomly determined and the manner in which they were grouped and spaced was governed by the expediencies involved i n synchronizing the taped script with the film passages. 54 The attributes used to characterize the Islanders were chosen at random from a subset of neutral t r a i t words in the Anderson (1968) l i s t . The items in the subset of t r a i t terms were selected for their appro-priateness as possible Orcadian descriptors by judges who had read the above-mentioned encyclopaedia essays on the Orkney Islands (see Appendix B, 1). c. Variations Two additional versions of the script were produced by modifying the primary text. In one variation i t was reported that the Islanders had very favourable concepts of Canadians. These longlasting impressions were said to have developed as a result of the economic benefits derived from Hudson's Bay ships which provisioned themselves at the islands, and also as a result of the frequent reports that Orcadians were being successful i n Canada (Appendix A, 2). Twenty positive attributes were included in the text as examples of "the types of terms that the Islanders use in characterizing Canadians" (Table 1). The words were selected from the traits in Anderson's (1968) l i s t which scored most highly on the "likableness dimension" and which f u l f i l l e d two additional requirements, one being that their antonyms were included in the set of lis t e d terms with the lowest likableness ratings, the other being that the words differ from each other with respect to their meaning (see Appendix B, 2). Table 1. The traits used to describe the Orkney Islanders i n the stimulus materials and the terms used to represent the Islanders' positive and negative characterizations of Canadians Traits attributed to the Orkney Islanders MODERATE AGGRESSIVE SELF-POSSESSED CRITICAL MEDITATIVE METICULOUS CONVENTIONAL UNGRACEFUL PROUD QUIET BLUNT IRRELIGIOUS SENTIMENTAL PERSISTENT DISCRIMINATING DAYDREAMER FEARLESS INOFFENSIVE NAIVE ECCENTRIC Traits used to represent the Islanders' characterizations of Canadians Positive SINCERE KIND BROAD-MINDED REASONABLE POLITE COOPERATIVE TOLERANT Negative INSINCERE UNKIND NARROW-MINDED UNREASONABLE IMPOLITE SELF-CENTRED INTOLERANT Positive ENERGETIC WISE FRIENDLY INTERESTING HUMOUROUS APPRECIATIVE RESPECTFUL Negative LAZY FOOLISH UNFRIENDLY UNINTERESTING HUMOURLESS UNAPPRECIATIVE DISRESPECTFUL Positive GOOD-TEMPERED AMUSING CLEAN-CUT GENEROUS FORGIVING BRIGHT Negative ILL-TEMPERED TIRESOME MESSY MEAN UNFORGIVING DULL aEXPORT-MINDED was also used unintentionally as an Orcadian descriptor. 56 Both the adjectival and the adverbial forms of the favourable t r a i t items were incorporated into the script and the otder of occurrence was randomly determined. A l l of the positive reciprocal stereotype information was embedded in the latter part of the documentary text where i t replaced redundant material from the basic script. The second variation on the primary text was similar i n form to the f i r s t . The twenty antonyms of the highly favourable words were presented as examples of "the types of terms in the Orcadians' negative stereotypes of Canadians" (see Table 1). They were included in random order, in the fin a l part of the script, i n both their adverbial and their adjectival forms. The Islanders' unfavourable characterization of Canadians in the negative reciprocal stereotype text was said to have resulted from the fact that the Hudson's Bay Company, in recruiting Orcadians for service in Canada, deprived the island communities of their much needed labour force, and having done so, did not treat the Orcadian emigrants as well as i t did other European recruits (Appendix A, 3). d. Presentation Each of the three passages was tape recorded over a background of traditional Gaelic folk tunes. The tapes were played in conjunction with the "silent" film. By stressing that the materials were f i r s t draft approximations to the f i n a l product i t was possible, i n the introductory talk, to provide a plausible reason for the unpolished nature of the documentaries. 57 2. Target groups As well as the Orkney Islanders, the Germans, Arabs, and Canadians were employed as target groups In the session designed to obtain the dependent stereotype measures. The choice of additional targets was based on the results of a preliminary study. Forty-four male and female undergraduates (enrolled in 200- and 300-level psychology courses) were asked to provide a complete l i s t of a l l of the ethnic, regional, and national groups that they could think of. They were then asked to rate each of the groups they had list e d on a nine-point favourableness scale. The tabulated responses indicated that the Germans and Arabs, respectively, were among the most salient groups in the neutral and negative ranges of the favourableness dimension. The findings also offered empirical support for the assumption that Canadians are a salient as well as a positive reference group for young student respondents (Appendix C). The inclusion of a neutral and an unfavourably perceived group served to control against the problems that can result from using a set of targets that i s imbalanced with respect to the social distance dimension (cf. Diab, 1962). More generally i t provided a broad context within which to measure the respondents' characterizations of the Orcadians. 58 Ques tionriaife materials 1. Post-film questionnaire Immediately following the presentation of each particular version of the documentary the respondents were asked to evaluate the stimulus mater-ia l s and to give their impressions of the Islanders on a 36-item question-naire (Appendix D). Some of the questions served as manipulation checks, others were f i l l e r items which were included to reinforce the cover story, and the remainder were intended to obtain information concerning the respondents' attitudes toward the Islanders and their impressions of the Islanders' demographic characteristics and general values. a. To check the effectiveness of the manipulations the respondents were asked whether or not Canadians were mentioned i n the film, whether or not they knew more about the Orkney Islanders after watching the film, how favourably (from very unfavourably to very favourably) they thought the Islanders were described in the film, and how favourably (on the same 5-item scale) they thought the Islanders view Canadians. b. In order to determine the details of their demographic 'profiles' of the Orkney Islanders the respondents were asked whether they thought the Orkney community is rural or urban, whether i t is economically developed or undeveloped, whether i t i s economically poorer than, richer than, or the same as Canada, whether the l i f e s t y l e i s the same as or different from that of Canadians, and whether the values of the Islanders are the same as or different from those of Canadians. They were asked also how 59 favourably they thought the Islanders see other people, how favourably they themselves view the Islanders, and whether they would like to v i s i t the islands, to live there for a long or a short period of time, or were just uninterested in the question. c. The items related to the cover story included questions such as: Do you think that the visual sequences should be changed? Do you think that a film of this sort should be LONGER, SHORTER, THE SAME LENGTH? Do you think that the soundtrack should be changed? If "YES", in what way? 2. Stereotype questionnaires a. Open-ended format The respondents were presented with an open-ended questionnaire in the f i r s t part of the stereotyping task (Appendix E). They were instructed to write down a complete l i s t of those traits and descriptive words that they f e l t were necessary to characterize adequately each of the four targets in question. Columns i n which to l i s t the responses were provided and additional directions to use a nine-point scale in order to rate the favourableness of each of the li s t e d attributes were also included. b. Check l i s t format The second part of the stereotyping task involved the use of a traditional, Katz and Braly-type check l i s t instrument (Appendix F). The respondents were asked to read carefully through a l i s t of seventy-five traits words and to tick those which seemed to them to be 60 typical of the target in. question. They were instructed further to write in additional words whenever they f e l t that these were necessary to 3 complete an adequate description of a particular group. Finally , they were asked to read over the set of traits that they had ticked and to rate each of the words on a nine-point favourableness scale. Space was provided on each of the questionnaire pages for the inclusion of those terms that the respondents might choose to add to the l i s t , and reminders were offered to ensure that a l l of the instructions had been completed. The seventy-five words in the adjectival check l i s t included those i n Table 1 as well as fifteen terms selected from the table of social stereo-types presented by Karlins, Coffman, and Walters (1969). The additional terms were: SCIENTIFICALLY-MINDED, TREACHEROUS, STRAIGHTFORWARD, MATERIALISTIC, EFFICIENT, STOLID, JOVIAL, SENSUAL, PHYSICALLY DIRTY, PLEASURE LOVING, AMBITIOUS, CRUEL, METHODICAL, EXTREMELY NATIONALISTIC, and DECEITFUL4. The l i s t was constructed by randomly ordering the t r a i t items'*, c. Information forms and other questionnaires 1. In both the film presentation and the stereotype measurement sessions each participant was given a "personal information" form which e l i c i t e d data concerning the respondent's age, date of birth, sex, occupation, nationality, birthplace, location of permanent residence, plans for future education, and whether or not he/she had lived outside the country for a long period of time (Appendix G). 61 2. As a part of the cover story for the documentary the respondents were told that the script was too lengthy and needed to be shortened. They were given a tabulation sheet on which to make notes about how much an what type of information they thought should be excluded from later ver-sions of the film (Appendix H). (The data yielded from this questionnaire were not of sufficient interest to warrant further attention.) 3. One further questionnaire served to ascertain whether or not the documentary presentation about the Orkney Islanders was the respondents' only source of information about the target group. At the end of the free response portion of the stereotyping task the participants were asked to l i s t a l l of the major sources of information that they had focussed upon when forming their impressions of each of the four targets^(Appendix I). Respondents Individuals from six groups of course cadets and three groups of officer candidates and course instructors at the National Sea Cadet Camp, H.M.C.S. Quadra, Comox, B.C., volunteered to participate in the study. The six groups of course cadets were allocated randomly to one of the three experimental conditions as were the groups of offi c e r candi-dates and course instructors. Thus each of the film tape-sequences was seen by two groups of younger course cadets and one group of older officer candidates and instructors. One hundred and ninety-three of the original 252 respondents success-fully completed a l l phases of the experiment. 62 Table 2. The location of permanent residence of the respondents in each of the experimental conditions and in the total sample CONDITION LOCATION B.C. PRAIRIES . ONTARIO QUEBEC MARITIMES POSITIVE FEEDBACK (n = 58) Percentage Number 29.3% 17 29.3% 17 15.5% 9 22.4% 13 3.4% 2 NO FEEDBACK (n = 67) Percentage Number 29.9% 20 25.4% 17 31.3% 21 10.4% 7 3.0% 2 NEGATIVE FEEDBACK (n - 68) Percentage Number 27.9% 19 23.5% 16 29.4% 20 14.7% 10 4.4% 3 TOTAL SAMPLE (« = 193) Percentage Number 29.0% 56 25.9% 50 25.9% 50 15.5% 30 3.6% 7 63 Table 3. The age distribution of respondents in each of the three experimental conditions and in the total sample CONDITION AGE CATEGORIES 14 years 15 years 16 years 17 years and younger * arid older POSITIVE RECIPROCAL CONDITION (n = 58) NON-RECIPROCAL CONDITION (n = 67) NEGATIVE RECIPROCAL CONDITION (n = 68) TOTAL SAMPLE (N = 193) Percentage 10.3% Number 6 Percentage Number 9.0% 6 Percentage 29.4% Number 20 Percentage 16.6% Number 32 34.5% 20 32.8% 22 33.8% 23 33.7% 65 25.9% 15 29.9% 20 14.7% 10 23.3% 45 29.3% 17 28.4% 19 22.0% 15 26.4% 51 64 Most of the 59 individuals who did not complete the task were respon-dents who decided not to attend the stereotyping session. In addition, however, a small number of people were advised to discontinue their pa r t i -cipation because they were French Canadians whose mastery of English was extremely limited. One respondent's data were excluded from the analysis because he reported that he had fallen asleep during the film. Two other sets were excluded because the participants indicated by their actions and by their responses that they had treated the study flippantly. Finally, eight individuals were called away from, or l e f t the stereotyping session before they had completed a l l of their tasks. Their responses were not included in the s t a t i s t i c a l analyses but, where appropriate, were employed In the construction of the social stereotypes (see Chapter 4: Scoring procedure and methods of analysis). Of the total sample of 193 respondents, 58 (fifty-four males and four females) saw the positive reciprocal stereotype documentary, 67 ( f i f t y - f i v e males and twelve females) saw the non-reciprocal documentary, and 68 (fifty-one males and seventeen females) were presented with the negative reciprocal stereotype film. Approximately 90% of the individuals in the sample were high-school students and the remainder were university undergraduates and skilled and semi-skilled workers. Their permanent homes were located throughout nine of the ten Canadian provinces. Prince Edward Island was the exception. British Columbia, Ontario, and Quebec were the most heavily represented provinces (see Table 2). 65 The respondents' ages ranged from 13 years and 7 months to 24 years and 11 months. The majority of the respondents, however, f e l l into the 15 (33.7%) and 16 (23.3%) year old age brackets (see Table 3). The median ages for the positive reciprocal, non-reciprocal, and negative reciprocal groups, respectively, were 16:4, 16:1, and 15:8; and the mean ages were 16.6, 16.6, and 16.0. A scrutiny of the mean and median ages and an examination of the age distributions in Table 3 suggest that the members of the negative reciprocal stereotype group were somewhat younger than their counterparts in the other experimental conditions. This observation was supported by an analysis of variance 7 (F(2,190) = 3.48, £ = .032). Because of this group d i f f e r -ence and because research has indicated that age i s an important variable in the stereotype formation process i t was necessary to examine the relationship between age and the dependent stereotype variables. The findings w i l l be discussed i n the next chapter. Setting The study was conducted at H.M.C.S. Quadra which i s a shore base located on a sandspit in the middle of Comox Harbour on Vancouver Island. The f a c i l i t y i s used as a summer camp for approximately 1,000 to 1,500 male and female Sea Cadets who spend two to ten weeks taking courses that can range from sailmaking to band music. Life at the camp i s vigorous and considerable emphasis i s placed on physical activities and general military discipline. Thus the experimental sessions which were held i n the late afternoon and evenings came at the end of f u l l and demanding days. 66 The locations as well as the. times of the experimental sessions were determined by the schedule of daily a c t i v i t i e s i n the camp. A l l of the documentary presentations were shown in the "OJT Lounge" just off the main parade square, while the stereotyping sessions were held in the OJT Lounge, in a nearby classroom, and i n a mess hall situated in a breeze-way between an administration building and a barracks. Procedure The procedure outlined in the following passages i s a modified version of one which was pretested i n a pilot study that involved 60 officer candidates from the Canadian Forces Officer Candidate School at C.F.B. Chilliwack, B.C. The nine groups which formed the sample of respondents were tested separately. The presentation of the film-tape sequence and the administration of the post-film questionnaire took approximately one hour. Each of the groups was recalled to the laboratory one week after the documentary presentation and the respondents were asked to complete the stereotyping task which required from one to two hours work. The experiment was conducted over a three week period. Phase 1 As the members of each particular group assembled in the "temporary laboratory" an officer requested their voluntary participation in the study and introduced the experimenter to them. He remained in the room u n t i l the experimenter had repeated the request for volunteers and read from a l i s t of "Basic Rights and Privileges of Volunteer Subjects " (Appendix J). 67 When those who wished to do so had l e f t , 116, the experimenter distributed a two-page handout which contained the cover-story for the f i r s t part of the study as well as a general description of the contents of the film-tape sequence. He allowed five minutes for the handout to be read and then reviewed the material verbally (Appendix K). The experimenter spoke about the "need" for the U.B.C. social science departments to develop their own documentary film for use in a number of courses and he outlined the form that these films would take: Our aim i s to give something of the flavour of the various life s t y l e s in different types of communities and cultures around the world. We hope to describe the characteristics of various  groups of people; to portray their l i f e s t y l e s and means  of earning a li v i n g ; to say something about how they view  other groups in the world around them; in general, to  outline their attitudes, customs and habits. A project to develop a pilot documentary about the Orkney Islanders was outlined, and the respondents were told that they would be required to assess some of the preliminary materials. The experimenter explained that by involving a wide range of individuals in the development of the film the fin a l product could be made not only more educationally useful but also more interesting. He went on to describe the nature of the scheduled presentation and he gave the respondents explicit instructions to attend to the content of the documentary script rather than the production features of the film. RATHER THAN ATTEND TO THE PRODUCTION ASPECTS OF THE FILM, IT IS BETTER IF YOU CONCENTRATED ON THE INFORMATION CONTAINED IN THE SCENES AND IN THE SOUNDTRACK. 68 TRY TO FORM A GENERAL IMPRESSION OF THE VISUAL SCENES  AND PAY CLOSE ATTENTION TO THE:SOUNDTRACK. THIS WILL HELP YOU IN YOUR ASSESSMENT OF THE FILM. LOOK FOR THE VARIOUS DESCRIPTIONS OF THE GENERAL CHARACTER OF THE ISLANDERS. NOTE THEIR PRINCIPAL OCCUPATIONS AND THE TYPES OF LINKS THEY HAVE WITH THE OUTSIDE WORLD. ALSO ATTEND TO THE INFORMATION THAT MAY GIVE YOU SOME IDEA OF HOW THEY VIEW VARIOUS "OUTSIDERS", THUS GETTING A GENERAL IDEA ABOUT THEIR ATTITUDE TO THE WORLD AND THE PEOPLE IN IT. DECIDE WHETHER THIS HAS ANY RELEVANCE TO YOU AS A CANADIAN. Following his introductory remarks the experimenter answered questions and distributed copies of a form (Appendix H) on which individuals could take notes about the items they f e l t should be excluded from the script. He then presented the film-tape sequence. When the documentary reached i t s mid-point the viewers were reminded briefly to keep their attention on the information i n the soundtrack and visual passages. Response booklets were distributed at the end of the film. They included a personal information form (Appendix G) and the post-film question-naire (Appendix D). Colour coded cards containing a code number were clipped to the front pages. When they had completed the "assessment task" the respondents were asked to: Please make sure that you have answered a l l of the questions, and that you have written your code number i n the appropriate space on the top of every page. Please hold on to the small card which l i s t s your code  number and bring i t with you when you come back next week. This w i l l save a great deal of time and effort. Since the current project has not yet been completed we would appreciate i t very much i f you do not discuss the details of the film script with friends and acquaintances. Thank you very much for your help and I hope to see you again next week. 69 Phase 2 When the respondents returned to the f i e l d laboratory at the end of a week they were once again reminded of their volunteer status. They were given an introduction handout which stated that they were about to p a r t i -cipate in a study to determine (1) how Canadians view a wide number of different social groups and (2) whether or not Canadians have ideas about how various nations see other people in the world (Appendix L). The experimenter elaborated upon the information in the handout and tried to make i t relevant to the experiences of the group members by using colloquial examples from l i f e at H.M.C.S. Quadra (Appendix M). The free response materials were distributed after the introductory remarks. The respondents were required to complete the tasks and return the forms to the experimenter in exchange for a booklet containing the check l i s t instruments. The experimenter outlined the response procedure and worked through an example of the open-ended task using an overhead projector. In addition he offered an "explanation" for the inclusion of the Orkney Islanders as a target that was intended to reduce the possibility of the respondents generating their own idiographic hypotheses about the g purpose of the testing session . There were five sections to the free response booklet. The f i r s t page contained a personal information questionnaire (Appendix G) and the second and third pages contained an example of the open-ended stereotyping task. Nine of the remaining ten pages included the instructions and response forms for the stereotyping tasks. On the f i n a l page the respondents 70 were asked to l i s t a l l of the sources of information, from which they had borrowed when forming their impressions of the Arabs, Canadians, Germans, and Orkney Islanders (Appendix I ) . The names of the four target groups were slisted in random order on the f i r s t four pages of the stereotype measurement section. The respondents were asked to describe each of the targets in their own words and to rate the terms they had used on a nine-point favourableness scale (Appendix E). The respondents were then asked to repeat the task from the perspective of the Orkney Islanders ( f i f t h page) and were presented with the names of the target groups which once again were ordered randomly (sixth through ninth pages). The instructions for the role playing portion of the task were as follows: Act as i f your identity was changed when you are doing these next few tasks. We would like you to assume a role as best you can. Carry out the following tasks AS IF YOU WERE AN ORKNEY  ISLANDER. Base your ratings on HOW YOU THINK YOU WOULD RESPOND IF  YOU HAD BEEN RAISED ON THE ORKNEY ISLANDS AND WERE LOOKING AT  THE WORLD THROUGH "ORKNEY ISLAND SPECTACLES". As the booklets were returned the responses were checked to ensure that none of the instructions had been overlooked. Following this procedure the check l i s t forms were administered, in turn, to subsets of four or five group members. The experimenter worked through the example in the booklet (Appendix N) to i l l u s t r a t e the procedure for completing the task, and the respondents were instructed to: 71 Read carefully through, the l i s t of t r a i t words and t i c k ( ) those which seem to you to be typical of If you do not find appropriate words for a l l of the typical characteristics you may WRITE-IN those which you think are necessary to complete an adequate description. Please read once more over the l i s t of traits that you have ticked ... and rate each of the words on the same scale that you used in the previous booklet: UNFAVOURABLE 1 2 3 4 , 5 6 7 8 9 FAVOURABLE j I I I I I ! I L Each set of materials included a two-page example, four randomized check l i s t forms on which the respondents were asked to characterize the target groups, instructions to assume the role of an Orkney Islander, and four additional, randomly ordered check l i s t forms on which the respondents were required to complete the reciprocal stereotyping task. When they had finished, the respondents were thanked for their co-operation and debriefed as carefully and as thoroughly as possible. It was not always easy to do this since a majority of the participants, although more than willing to work diligently on the various tasks, were not particularly interested In being enlightened about social psychological hypotheses and research hunches. Special problems associated with the data collection When I entered the " f i e l d laboratory" for the f i r s t time I discovered that many of the individuals were Francophones. This raised the immediate question of whether or not these potential respondents would understand the film script. It also raised the more long-term but equally important issue of whether or not they could complete the stereotyping task. As a f i r s t step toward resolving the problem I called for bilingual volunteers from each of the groups to act as interpreters. I asked them to 72 translate some i n i t i a l comments that I directed toward the French. Canadian cadets and also to translate the passages that I read from the various handouts. I then suggested that those individuals who did not think they could understand an English film script, or in fact, who could not understand me as I spoke, might prefer to opt out of the study. At the end of the session I spoke to the remaining Quebec respondents about whether or not they had understood the film. Those who said they had not were advised that they should think about whether or not they wanted to return the following week since a lot of English would also be involved in "that experiment". During the stereotype measurement session the interpreters once again translated the instructions that I read from the questionnaires. In addition I suggested that those who wished to do so could complete the tasks in French. Eighteen respondents accepted the offer. The decision to include the French responses in the data analyses was based on the Gardner, Kirby, Pablo, and Castillo (1975) finding that for bilingual respondents "the language of testing does not appreciably affect that product of stereotyping, or the stereotypes themselves" (p. 3), although i t does warrant careful consideration in multi-cultural research. The realization that English was a second language for some of the respondents was merely the f i r s t of a number of unexpected experiences during the study. I quickly realized just how demanding l i f e at the camp 73 was when a number of would be respondents promptly f e l l asleep as soon as the lights were dimmed and their compatriots settled down to watch the film. I can report with mixed feelings, however, that the number of sleepers in each group was considerably reduced by the tendency of the marching band to stage i t s best performances outside the windows of the "lab". These periodic episodes led to a number of unscheduled intermissions i n the film show. The respondents were not only remarkably good-natured. They were also very quick to respond to unusual stimuli. I found this out on a number of occasions when the f i r e alarm l e f t me with a room f u l l of half-finished questionnaires and the earnest hope that everyone could return before they were detailed for boat d r i l l or a two-day hike. Indeed, I was not so lucky on one occasion. I lost my group to a flash inspection of the barracks. It took me a week to recover these subjects, and this caused some methodological problems which w i l l be discussed i n the next chapter. The birthday party that threatened to s p i l l over into the mess h a l l where respondents were completing the stereotype questionnaires was no problem. If anything i t was a warm-up act to the grand finale of the study. At approximately ten-thirty on the last night, the female respondents in the last group were greeted with a row of "moons" as they characterized the target groups i n the breeze-way mess h a l l that was adjacent to a male barracks. In short i t was impossible to simulate the controlled conditions of a psychology department laboratory when running the study. Rather, the 74 stimulus materials were presented and the dependent data collected during brief l u l l s i n the comings and goings of the participants. Yet despite, or maybe because of these less than optimal experimental conditions the respondents remained attentive and interested, and responded diligently throughout the testing sessions. Scoring procedure and methods of analysis The eighteen sets of French responses were translated and then re-translated to ensure that the English items were equivalent to the French. Prior to scoring the data the participants' open-ended descriptions of each of the targets were compared with their check l i s t stereotypes. 9 The overlapping terms were deleted from the check l i s t characterizations (cf. Chapter 2). 1. Social stereotypes Social stereotypes of the Orkney Islanders were obtained from both the open-ended and check l i s t responses. In each particular instance they were constructed from the ten (or more) characteristics most frequently attributed to the target. As such they exemplify the types of terms in the core and periphery of the personal stereotypes of the members of the three experimental groups. Similar social stereotypes were generated to reflect the respondents' perceptions of how the Orkney Islanders charac-terize Canadians. 75 2. Personal stereotype variables Differentiation, favourableness, and commonness scores were derived from the core and peripheral elements of the respondents' personal stereo-types of the Orkney Islanders and also from the core and peripheral components of their estimations of the Orkney Islanders concepts of Canadians. Differentiation was measured by counting the number of different descriptive words in each characterization. Favourableness scores were computed for each respondents by averaging the favourableness ratings ascribed to the traits in each of the free response-.and check l i s t descriptions of the two t a r g e t s ^ (i.e., Orkney Islanders and Canadians as seen from an Orcadian's viewpoint). Measures of commonness were obtained from the core components of the personal stereotypes by comparing the terms in the open-ended descriptions with the most frequently used traits in the free response social stereo-types^. The total number of overlapping terms constituted the commonness scores. These scores could range from zero (no overlap with the social stereotype) to ten "(total overlap with the social stereotype). The procedure was repeated to derive commonness scores from the peripheral components of the personal stereotypes and from the components of the estimated reciprocal Canadian characterizations. A further measure involved computing the percentage of "common"  items in the core and periphery of each of the descriptions. 76 In addition, by calculating the number and percentage of the positive, neutral, and negative terms in both of the free response and check l i s t characterizations i t was possible to gain a broader perspective on the 12 differentiation and favourableness dimensions . Items scoring 7, 8, and 9 were considered to be positive t r a i t s ; those rated 4, 5, and 6 were categorized as being neutral; and the remaining terms which scored 1, 2, or 3 were considered to be negative characteristics. Two f i n a l sets of measures were derived from the core and periphery of the personal stereotypes of the Orkney Islanders. With respect to the stimulus materials i t was considered potentially informative to determine (1) how many of the traits in the film charac-terization of the Islanders appeared in the participants' personal stereo-types , and (2) to what extent.the terms used to detail the Orcadians' characterizations of Canadians were employed, in turn, by the respondents to describe the Islanders. Thus scores were tabulated for the individuals in each of the experimental groups which reflected the number of traits in common between their personal stereotypes of the Orkney Islanders and (a) the film characterization of the target group, (b) the positive "reciprocal characterization" of Canadians, and (c) the negative "reciprocal characterization" of Canadians. These stimulus overlap scores could range from zero (no overlap) to twenty (total overlap). A count was made of the number of positive, neutral, and negative traits that were common to the respondents' personal stereotypes of the Orkney Islanders and Canadians. By analyzing group differences i n these 77 content overlap scores i t was possible to test a general version of Campbell's C1967, p. 821) contrast hypothesis. Assuming that when the Orkney Islanders were described as disliking Canadians there was a greater contrast between them and the ingroup than when their attitude was unknown, and that, in turn, when their attitude toward Canadians was unknown there was a greater contrast than when the attitude was described as positive, we could expect that the content overlap between the personal stereotypes of Canadians and Orcadians would be least for the negative reciprocal group members and greatest for the individuals in the positive reciprocal experimental group. 3* Primary data analyses Chi-square analyses were employed to explore the relationships between experimental group membership and responses on the post-film questionnaire. Univariate, unweighted means, between-within analyses of variance were performed on the personal stereotype variables. (The decision to employ univariate rather than multivariate techniques was based on the fact that l i t t l e i s known about the robustness of the multivariate analogues of between-within ANOVAs in cases where the assumptions underlying the model are violated and the sample sizes are unequal). Following some preliminary methodological analyses (Ch. 5: Part 1), 3 X 2 between-within analyses of variance (see Figure 1 for design lay-out) were performed on the measures of differentiation, favourableness, common-ness, and 'percentage of common terms'. Experimental groups (3 levels) and stereotype components (2 levels) were the between-vand within- subjects independent variables, respectively. 78a T wo 3 X 2 X 3 , between-withiil ANOVAs were employed to examine the numbers and percentages of favourable, neutral, and unfavourable words in the personal stereotypes of the Orkney Islanders. In both of these cases the independent variables were: Experimental groups (3 levels), stereotype components (2 levels), and t r a i t categories (favourable, neutral, and unfavourable). Similarly, a 3 X 2 X 3 between-within analysis of variance was per-formed on the stimulus overlap scores. The three levels of the tr a i t  category factor represented the personal stereotype overlap with: (1) the positive reciprocal stimulus description of Canadians, (2) the stimulus material characterization of the Orkney Islanders, and (3) the negative reciprocal portrait of Canadians. The measures of content overlap between the personal stereotypes of the Orkney Islanders and those of Canadians were examined within the framework of a '3 (Experimental group) X 2 (Stereotype component) X 4 X 3 (Trait category: favourable, neutral, unfavourable) analysis of variance with repeated measures on the last three factors. The four-level factor included the (baseline) differentiation scores for the Orkney Islander stereotypes as well as the number of overlapping terms between the Orcadian personal stereotypes and those of the other three targets. Simple main effects and simple simple main effects analyses were conducted on significant two- and three-way interactions, respectively (Kirk, 1968; Winer, 1971), and post hoc pairwise comparisons were conducted using a t^  form suggested by Games (1971) 78b. 3 X 2 BETWEEN-WITHIN DESIGN Experimental groups  Positive reciprocal group Non-reciprocal A^ group Negative reciprocal A^ group Stereotype Component (Core) (Peripheral). B l B2 3 X 2 X 3 BETWEEN-WITHIN DESIGN Stereotype Component Experimental gr° uP s Positive reciprocal group (Core) B l Trait Category fav. neut unfav C l C2 C3 (Peripheral) Trait Category fav. neut unfav Non-reciprocal A£ group Negative reciprocal A_ group Figure 1. Design structure of the two- and three-way analyses of variance 79 that permits unequal ns and an. accurate, estimate of the stan-dard error of the difference between the two means. A signi-ficant outcome may be represented by ' t - X. - X. . . . -° 1 . 1 where CV i s the c r i t i c a l value for the q s t a t i s t i c . (Games, 1971, p. 101) Canonical correlation analyses, were employed to examine the relation-ships between the personal stereotype variables and the measures obtained from the respondents' estimates of the reciprocal characterizations of Canadians. The analyses were performed separately for each of the three experimental groups. Review The present chapter commenced with a sequential overview of the course of the study. It provided an outline of the stimulus materials and of the methods involved in their development. It contained a definition of the demographic characteristics of the respondents and a report on the procedure used to present the stimulus materials to them. The materials and procedure employed to measure the experimentally formed stereotypes were described, the dependent measures were defined, and the major types of data analyses were outlined. J2 80 CHAPTER FIVE: RESULTS AND DISCUSSION  1. Manipulation checks and the treatment of methodological problems Manipulation checks The potential success of the study depended not only on the respondents attending to the experimentally manipulated information in the film scripts, but also on them being unfamiliar with, and having no prior impressions of the Orkney Islanders. The questionnaire responses related to both of these issues are reassuring. One hundred and ninety-two of the one hundred and ninety-three respon-dents indicated that their concepts of the target group were based entirely 13 on the information in the stimulus materials . (The remaining experimental participant mentioned TV as a source of information but made no reference to the film.) Furthermore, the answers to the question regarding the favourableness of the Orkney Islanders' views of Canadians suggest that the respondents were aware of the reciprocal stereotype information embedded i n the documentary texts. Forty-seven (81%) of the individuals in the positive reciprocal condition f e l t that the Islanders view Canadians favourably. In contrast, forty (59%) of the negative reciprocal group members considered that Canadians are viewed in unfavourable terms by the Orcadians (Table 4). It i s interesting to note with reference to the responses i n Table 4 that the individuals who were not presented with reciprocal information Table 4. The number and percentage of the individuals in each of the experimental groups and i n the total sample choosing each of the possible responses to the post-film questions pertaining to the experimental manipulations Question Response Frequency Positive Non- Negative Total Chi-Square (para- Categories and % a of reciprocal reciprocal reciprocal Sample Values b phrased) response group group group (N=193) (n=58) (n=67) (n=68) How do Favourably n 47 24 8 79 the % 81.0% 35.8% 11.8% 40.9% Orkney Neutrally n 1 25 10 36 Islanders % 1.7% ,37.3% 14.7% 18.7% view Canadians? Unfavourably n 5 2 40 47 % 8.6% 3.0% 58.8% 24.4?. No response n 5 16 10 31 % 8.6% 23.9% 14.7% 16.1% * = 108.86, df = 4, £ <.0005 Table 4 (continued) Question Response Frequency (para- Categories and % a of phrased) responses Positive Non- Negative reciprocal reciprocal reciprocal group group group (n=58) (n=67) (n=68) Total Sample (N=193) Chi-Square Values'3 How do Favourably n 33 18 11 62 the % 56.9% 26.9% 16.2% 32.1% Orkney Neutrally n 15 29 13 57 X = df = 61.00, 4, Islanders % 25.9% 43.3% 19.1% 29.5% £< .0005 see others? Unfavourably n 4 7 38 49 % 6.9% 10.5% 55.9% 25.4% No responses n 6 13 6 25 % 10.3% 19.4% 8.8% 13.0% Does the YES n 37 26 44 107 film % 63.8% 38.8% 64.7% 55.4% mention NO 14 X = 13.56, n 7 8 29 df = 4, Canadians? % 12.1% 20.9% 11.8% 15.0% P_ < .009 CAN'T SAY n 13 26 12 51 % 22 A% 38.8% 17.6% 26.4% No response n 1 1 4 6 % 1.7% 1.5% 5.9% 3.1% Table 4 (continued) Question Response Frequency Positive Non- Negative Total Chi-Square (para- Categories and % of reciprocal reciprocal reciprocal Sample Values phrased) responses group group group (N=193) (n=58) (n=67) (n=68) How Favourably n 43 37 32 112 favourably % 74.1% 55.2% 47.1% 58.0% • i_ were the * = 16.33, Neutrally n 11 19 18 48 df - 4, Orkney % 19.0% 28.4% 26.5% 24.9% £ < .003 Islanders described? Unfavourably n 1 3 12 16 % 1.7% 4.5% 17.6% 8.3% No response n 3 8 6 17 % 5.2% 11.9% 8.8% 8.8% Percentages are column percentages. Only the members of each of the three experimental groups who provided definite responses were included in the analyses. 84 about how Canadians are perceived by the Islanders were evenly divided between those who f e l t that Canadians are viewed favourably (24) and those who f e l t that Canadians are viewed in neutral terms(25). Only 2 of the 51 group members who answered the question considered that the Islanders see Canadians unfavourably. Additional evidence that the participants attended to the relevant sections of the documentary passage is provided by the responses to the question concerning whether or not Canadians were referred to in the film. Individuals in both the positive and negative reciprocal conditions were much more lik e l y than those in the non-reciprocal group to agree that 2 Canadians had been mentioned ( = 13.56, df= 4, p_<.009, see Table 4 for details of the response frequencies). The respondents appear to have generalized from their perceptions of how the Orkney Islanders characterize Canadians to their impressions of how the Islanders view others in general. There is a significant correlation between scores on both variables (r= 0.370, p_<.01). Indeed thirty-three (57%) of those in the positive reciprocal condition believed that the Orcadians view others favourably, twenty-nine (43%) of the individuals in the non-reciprocal group f e l t that others are seen neutrally by the Islanders, and thirty-eight (56%) of the participants in the negative reciprocal condition indicated that the Orcadians view others in unfavourable terms (see Table 4). In brief the experimentally manipulated information formed a salient portion of the documentary passages, i t shaped the respondents' perceptions 85 of how the Orkney Islanders view Canadians, and i t influenced their impressions of the Islanders' more general outlook toward others. On the other hand the individuals' ratings of the favourableness of the film description of the Orkney Islanders did not mirror the neutral characterization that was embedded in the stimulus materials. In fact a major portion of the group members in each of the experimental conditions indicated that the Islanders were favourably described (Table 4). However.while 74% of the positive reciprocal group members f e l t that the Orcadians were characterized in favourable terms only 47% of the individuals in the negative reciprocal group made the same judgement. Moreover, 18% of those in the negative reciprocal condition f e l t that the Islanders were described unfavourably while only one person in the positive reciprocal group thought so. Thus there is a significant difference between groups 2 1 with respect to the pattern of responses (7<- = 16.33, df= 4, p_= .003) and this difference appears to be due to a "halo effect" resulting from the manner in which the Orkney Islanders were portrayed as viewing Canadians. Pretest on stereotype data from the negative reciprocal group One of the subgroups of course cadets i n the negative reciprocal condition was unable to participate in the stereotype measurement session unt i l two weeks after the film presentation. Consequently before pooling the data i t was necessary to establish whether or not the responses from this group were comparable to those from the other negative reciprocal subgroups. 86 A test of time-lag effects was conducted by comparing the responses of the "two week test-interval" group members (n= 19) with those of the individuals in the "one week test-interval" course cadet subgroup"^ (n= 33). Scores on the differentiation, favourableness, commonness, and 'percentage of common terms' variables were analyzed using 2 (Group) X 2 (Stereotype component) between-within, unweighted means analyses of variance. None of the group main effects nor group X stereotype component interactions was significant (with df= 1,50, p_>.12 i n a l l cases but one, the exception having a p_ value of .09). Similarly the analyses on both the percentage and number of positive, neutral, and negative terms, and on the stimulus overlap scores were conducted with 2 (Group) X 2 (Stereo-type component) X 3 (Level of measurement variable) between-within, unweighted means analyses of variance. Once again, none of the between-group differences nor group X stereotype measure interactions was signi-ficant ( a l l £sy .13). Finally, the main effects and group X repeated measure interactions in the analysis on the content overlap scores were non-significant (p_> .11 in every case). Since there was no significant difference between groups on any of the variables i t was considered appropriate to Include the data from the "test-delayed" subgroup in the major analyses. Homogeneity of variance pretests The work of Petrinovich and Hardyck (1969) draws attention to the importance of testing the homogeneity of variance assumption for ANOVA designs with unequal sample sizes, since: 87 The combination of unequal sample size and unequal variance which produced drastically incorrect values of t^  ... has a similar effect on a l l multiple comparison methods, (p. 47) Indeed the use of percentage-score variables offered particular cause for concern in the present study since percentage scales "generally do not provide homogeneity of variance (Winer, 1971, p. 537)." Hartley's I,^ 1 5 (cf• Kirk, 1968, p. 62) was employed to test the assumption of homogeneity of variance for each of the main effects and interactions (simple or simple simple main effects) on each of the personal stereotype variables. The favourableness scores were the only measures to conform to the assumption. The analyses on a l l other variables yielded significant F values denoting unequal within-cell variances. Moreover, -max in a majority of cases the largest variances were associated with the smallest samples, causing the probability of a type 1 error to be greater than o\ (Glass and Stanley, 1970, p.372). Thus i t was necessary to interpret the results of the s t a t i s t i c a l analyses on the dependent variables with caution. The repeated measures analyses on a l l of the personal stereotype variables, except the favourableness scores, were conducted using Geisser-Greenhouse conservative F_ tests. These approximate F_ tests offer protection against the increased probability of type 1 error that i s associated with repeated measures J_ tests i n which the homogeneity of variance and covariance assumptions*^ are violated. Type 1 error i s guarded against by employing reduced degrees of freedom. For example, in a one-way, k treatment, repeated measures, conservative analysis of variance the c r i t i c a l value i s F (.1, n - 1) (cf. Kirk, 1968, p. 142-143; Winer, 1971, p. 526). 88 Other than employing different degrees of freedom with, which, to establish the c r i t i c a l values of the J_ ratios the computational procedures for the conventional and conservative F_ tests are identical (cf. Winer, 1971, p. 542). Correlations between age and stereotype Variables Because of the significant difference between groups with respect to the age of the respondents, correlations between measures of age (in years) and each of the personal stereotype and perceived reciprocal stereo-type variables were calculated. The data from each of the three groups were treated separately. The correlations between age and the core personal stereotype differentiation scores, and age and the number of neutral words in the core components of the personal stereotypes were significant. When averaged across groups the rs were .295 in each case (E<.05). In general terms, the older the respondents were the more items they had in their free response descriptions of the Orkney Islanders. Thus i t was necessary to consider this finding when discussing the between-groups analysis on the personal stereotype differentiation scores. Summary of methodological results Questionnaire responses indicated that the experimentally manipulated information formed a salient portion of the documentary passages. It was attended to by the respondents, and i t influenced not only the participants' perceptions of the Orkney Islanders' attitudes toward Canadians but also their impressions of the Islanders' more general outlook toward others. 89 Contrary to expectations, however, a major portion of the individuals i n each of the experimental groups f e l t that the Orcadians were described in favourable rather than in neutral terms. The tendency for respondents to make this observation was greatest in the positive reciprocal condition and least evident among the members of the negative reciprocal group. Comparisons between the responses of the course cadets in the negative reciprocal condition whose stereotypes were eli c i t e d two-weeks after the film presentation and those of the 'negative reciprocal' course cadets whose stereotypes were measured after the standard one-week interval yielded no inter-group differences. Thus i t was possible to include the data from the "retest delayed" subgroup in the major analyses. Analyses indicated that the homogeneity of variance assumptions were violated on a l l variables except the favourableness scores. Moreover, the largest variances were generally associated with the smallest samples. Consequently, to protect against type 1 error, Geisser-Greenhouse conser- : vative F tests were employed for a l l of the repeated measures ANOVAs on the personal stereotype variables. Finally, i t was determined that there was a significant positive correlation between age and the degree of differentiation of the core components of the personal stereotypes. This finding was important with respect to the discussion of the analysis of the personal stereotype differentiation scores. 90 2. Presentation and discussion of experimental results  a. Post-film impressions Of the Orkney Islanders and estimates of the Islanders' reciprocal stereotypes of Canadians The analyses of the respondents' post-film impressions of the Orkney Islanders and of their estimates of the Islanders' reciprocal stereotypes of Canadians serve to elaborate upon the manipulation checksfindings and to provide a more detailed account of the manner in which the members of the different experimental groups construed the stimulus materials. As a consequence i t becomes possible to discuss the personal stereotype results not only with reference to the content of the documentary presentations but also with reference to the respondents' own recollection of the film information. Post-film impressions of the Orkney Islanders Table 5 presents the modal responses of the positive, negative, and non-reciprocal group members to the questions concerning their post-film impressions of the Orkney Islanders. (More comprehensive tables of responses are included in Appendix 0.) A majority of the individuals in each of the experimental groups characterized the Islanders as a rural people who are poorer than . Canadians. However, there was somewhat less agreement among the respondents concerning their views about the Orcadians1 level of economic development and the nature of the Islanders' values and l i f e s t y l e . A l i t t l e more than half of the negative and non-reciprocal group members who answered the pertinent question rated the Islanders as being underdeveloped (see Appendix 0). In contrast, the Orcadians were judged Table 5. Modal responses of positive, negative, and non-reciprocal group members to questions concerning their post-film impressions of the target Positive reciprocal group (n=58) Frequency Modal and % of response response n % Non-reciprocal group (n=67) Frequency Modal and % of response response n % Negative reciprocal group (n*°68) Frequency Modal and % of Chi-square Values response response n % The Orkney Islanders are: RURAL 44 ECONOMICALLY DEVELOPED and 'ZW.&> POORER than Canadians 36 They have: and the DIFFERENT VALUES SAME LIFESTYLE as Canadians 7?6% 62% RURAL 46 ECONOMICALLY UNDER-DEVELOPED 31 36 62% and POORER than 49 Canadians 69% RURAL ECONOMICALLY 46% UNDER-DEVELOPED and 73% POORER than Canadians DIFFERENT 18 31% VALUES and a DIFFERENT DIFFERENT 31 46% VALUES and a DIFFERENT 26 45% LIFESTYLE 33 49% LIFESTYLE than Canadians than Canadians 41 60% *X= 2.37, df=2, ns 31 46% X= 4.19, df=2, ns 49 72% 7(= 6.52, df=4, ns ns 23 34% 3.59, df=2, 45 66% *X=11.36, df=2, p<.005 The complete tables on which the chi-square analyses were performed are listed in Appendix 0 bThe same number of respondents (18) f e l t that the Islanders' values were the SAME as Canadians. Thus the respondents in this condition were equally s p l i t with respect to their impressions of the similarities between Canadian and Orcadian values 92 to be economically developed by approximately two-thirds of the respondents (n=36). in the positive reciprocal condition. A similar contrast between the positive reciprocal group members and the individuals in the other conditions occurred with respect to the remaining questions. While the positive reciprocal respondents were equally divided on the issue of whether or not the Orcadians' values are different than Canadians', and were inclined generally to characterize the Islanders' l i f e s t y l e as being the same as that of their own national group, a major portion of the Individuals i n the non-reciprocal and negative reciprocal conditions considered the values and l i f e s t y l e of the Islanders to be different thahfethose of Canadians. Although the findings do not reflect substantial differences between groups, and indeed only achieve significance with respect to the judgements of the Islanders' l i f e s t y l e 11.36, df_= 2, p_4.005), they do suggest a tendency on the part of the individuals i n the negative and non-reciprocal conditions to diffe r from the positive reciprocal group members regarding their interpretation of the stimulus information about the demographic characteristics of the Islanders. It appears that when a combination of attitudinal and demographic information i s presented about a target, the attitudinal information may influence people's evaluations of the target's distinguishing demographic features. This possibility has implications for our stereotype formation model, and these implications w i l l be discussed in a later chapter. Table 6. Most frequently occurring items i n the respondents' estimates of the Orkney Islanders' stereotypes of Canadians CORE COMPONENTS (free response format) PERIPHERAL COMPONENTS (check l i s t format) % of group Number of Average % of group Number of Average TERM checking group favour- TERM checking group favour-term members ableness term members ableness Canadians are: (sample n =61)a checking term ratings Canadians are: (sample n =60)a checking term ratings FRIENDLY 43% 26 8.3 GOOD-TEMPERED 52% 31 7.2 HELPFUL 20% 12 8.8 PROUD 48% 29 8.1 RICH 18% 11 7.5 WISE 47% 28 7.8 KIND 18% 11 7.7 REASONABLE 47% 28 7.6 WEALTHY 15% 9 7.8 GENEROUS 47% 28 7.7 PROUD 13% 8 8.1 SINCERE 47% 28 7.7 INTELLIGENT 13% 8 7.0 COOPERATIVE 45% 27 7.9 GENEROUS 13% 8 8.1 ENERGETIC 45% 27 7.6 COOPERATIVE 11% 7 7.4 SCIENTIFICALLY-MINDED 45% 27 7.8 GREEDY 10% 6 4.3 AMUSING 43% 26 6.9 oo IT, 1 p-l u 3 H t—I SA o tu Table 6 (continued) CORE COMPONENTS (free response format) PERIPHERAL COMPONENTS (check l i s t format) TERM Canadians are: % of group Number of Average % of group Number of Average checking term group members (sample n checking »70) a term favour-ableness ratings TERM Canadians are: checking term group members (sample n checking =67)a term favour-ableness ratings SI h cu P g o 8 tu M U 8 FRIENDLY 34% 24 7.8 SCIENTIFICALLY-MINDED 48% 32 7.5 RICH 20% 14 6.9 POLITE 48% 32 7.4 KIND 16% 11 7.7 AMUSING 46% 31 6.9 WEALTHY 16% 11 6.2 GENEROUS 45% 30 7.2 SMART 13% 9 7.7 FORGIVING 45% 30 7.0 HELPFUL 13% 9 7.3 REASONABLE 45% 30 7.2 PEACEFUL 13% 9 7.6 PROUD 45% 30 7.7 NICE 11% 8 7.6 ENERGETIC 40% 27 6.5 HEALTHY 11% 8 7.2 COOPERATIVE 40% 27 7.7 HAPPY 11% 8 7.9 KIND 37% 25 7.5 Table 6 (continued) CORE COMPONENTS (free response format) PERIPHERAL COMPONENTS (check l i s t format) % of group Number of Average % of group Number of Average TERM checking group favour-.. TERM checking group favour-term members ableness term members ablenes Canadians are: (sample n checking ratings Canadians are: (sample n checking ratings =70) a term =69)a term RICH 33% 23 6.9 SELF-CENTRED 43% 30 3.6 FRIENDLY 16% 11 7.5 PROUD 43% 30 7.7 SELFISH 14% 10 1.6 UNFRIENDLY 41% 28 2.6 GREEDY 10% 7 1.7 FOOLISH 39% 27 3.1 UNFRIENDLY 10% 7 2.3 SCIENTIFICALLY- 36% 25 7.3 MINDED LAZY 10% 7 1.4 LAZY 36% 25 3.0 INTELLIGENT 9% 6 8.2 ILL-TEMPERED 33% 23 2.6 STUPID 7% 5 1.3 PLEASURELOVING 33% 23 6.5 KIND 7% 5 6.2 AMBITIOUS 33% 23 6.8 (medium)b SMART 7% 5 6.6 NARROW-MINDED 32% 22 2.8 PROUDC 7% 5 6.5 VAINC 7% 5 1.2 WEALTHY0 7% 5 7.2 oo I vol § M U 3 S3 o W 53 vO Table 6 (continued) ^ a t a from the eight incomplete sets of responses (see Chapter Four: Respondents) were used, where possible, in compiling the social stereotypes. The increased sample sizes are noted in the percentage column headings, while the basic sample sizes of the experimental groups are listed in the vertical spanner headings. bOne of the respondents used this term to modify the t r a i t word C T r a i t s not used as comparison items in the computation of the commonness scores vO O N 97 Respondents' estimates of the Orkney Islanders' stereotypes of Canadians  Contents The terms which occurred most frequently in the respondents' free response and check l i s t estimates of the Orkney Islanders'reciprocal characterizations of Canadians are listed in the social stereotypes in Table 6. The contents of the positive and non-reciprocal group members' reciprocal social stereotypes are highly favourable. Indeed there i s substantial overlap between the traits in both portraits, particularly among the peripheral components. There are favourable t r a i t s , too, in the social stereotype of Canadians derived from the negative reciprocal respondents' role-played characterizations. Among these are such terms as "friendly," "rich," "kind," "proud," and "scientifically-minded," which were commonly used by the individuals in each of the three experimental groups. Nevertheless, despite the inclusion of these positive words, the negative reciprocal profile i s by and large an extremely unfavourable one. Generally speaking, then, the contents and evaluative tone of the reciprocal social stereotypes Indicate that the positive and negative reciprocal group members remained aware of the Orkney Islanders' supposed views about Canadians during the week that intervened between the presentation of the documentary information and the e l i c i t a t i o n of the stereotype data. They also indicate that even in the absence of feedback about the Islanders' stereotypes of Canadians, the non-reciprocal respondents readily inferred what these characterizations might be. 98 Table 7. Means and standard deviations for the differentiation and favourableness scores for the core and peripheral components of the respondents' estimates of the Orkney Islanders* reciprocal stereotypes of Canadians Stereotype Component Positive reciprocal group (n=58) Non-reciprocal group (n=67) Negative reciprocal group (n=68) c Q> U u M-l Core M S D 5.4 4 > Q Periphery M S D 1 5 , 9 7 1 5.5 14.6 3.1 5.9 4.6 15.7 1.9 8.9 to o o to to to <u C i 0) H ! JO « Vi 3 I Core M SD 6 * 8 1.9 Periphery M g D 6 * 8 1.5 6.3 6.4 1.6 1.5 4.4 4.9 2.2 1.9 99 Structure Additional information about the characteristics of the respondents' estimated reciprocal stereotypes is provided by the dependent measures summarized in Table 7 and in Appendix P. Table 7 contains the mean differentiation and favourableness scores for the core and peripheral components of the estimated reciprocal characterizations. The means and standard deviations for the remaining variables (e.g., "commonness" and "stimulus overlap") are lis t e d in Tables A, B, and C in Appendix P. These data, when reviewed in conjunction with the information in Tables 4 and 6, provide an answer to the f i r s t of the model-testing questions: Will respondents make inferences about the Orkney Islanders' stereotypes of their own national group In the absence of definite information about the nature of these "reciprocal stereotypes"? The evidence suggests that such inferences were made by the individuals in the non-reciprocal condition. A l l of the sixty-seven members of the non-reciprocal group completed the reciprocal stereotyping task, characterizing the four stimulus groups from the Orkney Islanders' perspective. In so doing they produced estimates of the Orcadians' reciprocal stereotypes of Canadians that were as differentiated* 7 as those provided by the individuals in the positive and negative reciprocal experimental conditions (Table 7). Furthermore, a scrutiny of the percentage scores in Table 6 (and the commonness scores in Table A, Appendix P) indicates that when estimating which particular traits the Orkney Islanders are most l i k e l y to attribute to Canadians, the interpersonal agreement between respondents in the 100 non-reciprocal group was approximately the same as that between 18 individuals in each of the other experimental conditions. The summary data In the above mentioned tables also contain evidence related to the f i r s t part of the second model-testing question: Will information concerning the Orkney Islanders' reciprocal stereotypes of Canadians be attended to when i t i s included in a documentary script about the Islanders? The post-film questionnaire responses (Table 4) and the reciprocal stereotype characterizations (summarized i n Tables 6 and 7, and in Tables A, B, and C, Appendix P). indicate that the individuals In the positive and negative reciprocal conditions did pay attention to the relevant information about the manner in which the Islanders were said to characterize Canadians. They correctly judged the favourableness of the reciprocal stereotypes that were embedded i n the film presentations and, in addition, they reproduced much of the content of these characterizations in their own (check l i s t ) estimates of how the Orcadians describe Canadians. For example, immediately after viewing the stimulus materials the respondents in the positive reciprocal condition agreed that Canadians are seen favourably by the Orkney Islanders. Their perceptions were articulated in greater detail a week later when they performed the reciprocal stereotyping task. Over sixty percent of the items in their role-played characterizations of Canadians were favourable (Table B, Appendix P). Many of these, particularly the check l i s t terms that were used most frequently by the group members, were words that had been included in the "positive reciprocal documentary portrait" of Canadians c 101 (see Table 6 and Table C, Appendix P). Similarly, a large majority of the negative reciprocal group members affirmed that Canadians are viewed unfavourably by the Orkney Islanders (Table 4). Their estimates of the Orcadians' reciprocal stereotypes of Canadians, particularly the core components, contained a high percentage of undesirable terms (Table B, Appendix P), and they included, among their peripheral components, many of the traits that were employed to describe Canadians in the negative reciprocal stimulus script (Table 6 and Table C, Appendix P). In summary, the non-reciprocal group members made inferences about the Orkney Islanders' perceptions of Canadians in the absence of definite cues about the nature of these reciprocal stereotypes. Furthermore, the respondents in the positive and negative reciprocal conditions attended to the reciprocal stereotype information embedded i n the stimulus materials. They made correct judgements about the evaluative tone of the information, and they employed many of the traits comprising the reciprocal information when they estimated the Orkney Islanders' characterizations of Canadians. Moreover, the estimates of the Orkney Islanders' reciprocal stereotypes of Canadians that were provided by the individuals in the non-reciprocal condition were similar in content and structure to those provided by the members of the positive reciprocal group: A finding which i s i n keeping with the frequently observed preference among respondents for evaluating and perceiving people positively (e.g., Johnson and Downing, 1976; Zajonc and Burnstein, 1965). 102 Overview Before proceeding to discuss the personal stereotype data i t i s useful to collate the findings regarding the respondents' construal of the stimulus materials. The positive reciprocal group members saw the Orkney Islanders as a rural, although economically developed people who are poorer than Canadians, who have different values but the same l i f e s t y l e as Canadians, and who consider Canadians to be friendly, good-tempered, helpful, and proud. The negative reciprocal respondents, on the other hand, perceived the Orkney Islanders to be rural and economically underdeveloped, to be poorer than Canadians, to have different values and a different l i f e -style than Canadians, and to characterize Canadians as being rich, s e l f -centred, proud, unfriendly, and foolish. The individuals i n the non-reciprocal condition viewed the Islanders as a rural, underdeveloped people who are poorer than Canadians, who have different values and a different l i f e s t y l e than Canadians, and who describe Canadians as being friendly, rich, kind, scientifically-minded, and polite. b. The contents and structure of the respondents' personal stereotypes of the Orkney Islanders Contents Table 8 contains the l i s t s of traits most frequently attributed 19 to the Orkney Islanders by the members of the three experimental groups. The contents of these social stereotypes are predominantly favourable, and words such as "friendly," "hardworking," "co-operative," "happy," "proud," and "kind" were among those most commonly used by the individuals in each of the positive, negative, and non-reciprocal treatment conditions. The percentage values i n Table 8 reflect the level of Interpersonal consensus i n the assignment of social stereotype traits to the Islanders. Table 8. Social stereotypes of the Orkney Islanders constructed from the most frequently occurring items in the personal stereotypes of the respondents in the three experimental groups CORE COMPONENTS (free response format) PERIPHERAL COMPONENTS (check l i s t format) TERM Orkney Islanders are: % of group checking term (sample n =61)a Number of group members checking term Average favour-ableness ratings TERM Orkney Islanders are: % of group checking term (sample n =60)a Number of group members checking term Average favour-ableness ratings co Pi l=> § pi M O w > H H M w O Pi FRIENDLY POOR SIMPLE HARDWORKING PEACEFUL HAPPY PROUD RICH SOCIAL/SOCIABLE* COOPERATIVE NICEC 36% 31% 21% 16% 15% 15% 15% 11% 10% 8% 8% 22 19 13 10 9 9 9 7 6 5 5 8.0 4.6 6.4 8.1 6.4 7.3 6.9 5.0 6.3 5.6 7.8 COOPERATIVE PROUD GENEROUS REASONABLE KIND GOOD-TEMPERED POLITE SINCERE RESPECTFUL AMUSING QUIETC STRAIGHTFORWARD1 CLEAN CUTC FRIENDLY0 68% 67% 63% 60% 58% 58% 57% 53% 52% 50% 50% 50% 50% 50% 41 40 38 36 35 35 34 32 31 30 30 30 30 30 7.4 8.1 7.5 7.2 8.1 7.4 7.4 7.6 7.9 7.1 7.1 7.4 7.0 7.4 Table 8 (continued) CORE COMPONENTS (free response format) PERIPHERAL COMPONENTS (check l i s t format) % of group Number of Average % of group Number of Average TERM checking group favour- TERM checking group favour-term members ableness term members ableness Orkney Islanders (sample n checking ratings Orkney Islanders (sample n checking ratings are: =70)a term are: =67)a term FRIENDLY 39% 27 8.1 COOPERATIVE 70% 47 7.2 POOR 29% 20 4.6 SINCERE 67% 45 7.3 PEACEFUL 17% 12 8.2 POLITE 61% 41 7.1 KIND 16% 11 8.0 PROUD 60% 40 7.4 HARDWORKING 14% 10 7.6 QUIET 57% 38 5.6 QUIET 14% 10 5.8 GENEROUS 49% 33 7.1 OLD-FASHIONED 13% 9 5.1 GOOD-TEMPERED 49% 33 7.4 HAPPY 13% 9 8.2 ENERGETIC 48% 32 7.8 NICE (PEOPLE/ 13% 9 8.0 SENTIMENTAL 48% 32 6.1 PERSONALITY ) ISOLATED 11% 8 5.3 FRIENDLY 48% 32 7.9 sot M u i l KIND 48% 32 7.1 Table 8 (continued) CORE COMPONENTS (free response format) PERIPHERAL COMPONENTS (check l i s t format) % of group Number of Average % of group Number of Average TERM checking group favour- TERM checking group favour-term members ableness term members ableness Orkney Islanders (sample n_ checking ratings Orkney Islanders (sample n checking ratings are: =70) a term are: =69) term (VERY) FRIENDLY/ 31% 22 7.3 PROUD 61% 42 7.8 FRIENDS (SEMI-) POOR 23% 16 3.6 POLITE 48% 33 7.6 HARDWORKING 19% 13 8.0 COOPERATIVE 45% 31 7.7 OLD-FASHIONED 19% 13 5.4 REASONABLE 45% 31 7.3 HAPPY 17% 12 8.0 QUIET 42% 29 5.6 QUIET 10% 7 4.9 ENERGETIC 41% 28 7.7 KIND 10% 7 7.6 CLEAN-CUT 41% 28 7.5 PREJUDICED 10% 7 1.6 STRAIGHTFORWARD 39% 27 6.6 ISOLATED 9% 6 5.5 SINCERE 38% 26 7.7 RICH 9% 6 5.7 SELF-CENTRED 36% 25 4.4 PROUD° 9% 6 7.7 oo V C 8 CD o 8 fu H U > H ^ Table 8 (continued) "Data from the eight incomplete sets of responses (see Chapter Four: Respondents) were used, where possible, in compiling the social stereotypes. The increased sample sizes are noted in the percentage column headings, while the basis sample sizes of the experimental groups are listed in the vertical spanner headings bWords with similar or overlapping meaning were grouped to better represent the frequency with which the particular semantic categories were employed. The most commonly used items are presented f i r s t Traits not used as comparison items i n the computation of the commonness scores "Swords in parentheses were used as t r a i t modifiers by at least one of the respondents 107 Table 9. Mean commonness and 'percentage of common terms' scores for the core and peripheral components of the personal stereotypes of the Orkney Islanders' Stereotype Component co o CO 01 u e p. > O) 4-1 .g 4J o 3 e u c o te e CO o 0> o c JE •rl Core Periphery Positive reciprocal group (n=58) 1.8 6.0 Non-reciprocal group (n=67) 1.8 5.6 Negative reciprocal group (n=68) 1.6 4.4 01 co co to a 01 id C p. 4J 0) >. c 4-1 4-1 0) o u 6 01 o (-1 0) 0) 4-1 o to c V CO e s o •H Core Periphery 30.3% 35.8% 32.4% 38.7% 30.7% 29.8% 108 Table 10. Analyses of variance of the commonness and 'percentage of common terms' scores Commonness % commonness Source d f a MS F MS F Between Ss ** 8.92 Experimental group (A) 2 31.25 966.43 2.96 Error 190 3.50 326.49 Within Ss Stereotype component (B) 1 1230.39 ** 377.14* 1311.49 * 4.43 A X B 2 15.50 * 4.75 532.50 1.80 Error 190 3.26 295.84 ^ i t h only two levels of the within-subjects variable the degrees of freedom for the Conservative J_ tests are the same as those for the conventional tests (cf. Winer, 1971, p.526) *£<.05 109 In turn, the commonness and "percentage of common terms" scores indicate the extent to which, the items in each of the personal stereotypes are represented in the social stereotypes. Thus they provide further information about the degree of uniformity between the contents of the respondents' characterizations. The means for both of these variables are reported in Table 9. The analysis of the commonness scores yielded a significant inter-action between experimental conditions and stereotype components (Table 10). Tests of the simple main effects confirmed what i s apparent from a scrutiny of the means in Table 9: namely, that there was no significant difference between groups with respect to the number of "common" terms in the core elements of the personal stereotypes (F (2,380)<1), and that the peripheral elements of the respondents' characterizations included a greater number of commonly used t r a i t words than did the core components (p ^.001 for each of the treatment conditions), ^cons ^ The simple main effects tests also indicated that there was a significant difference between treatment groups with respect to the number of "common" words in the individuals' peripheral stereotype components (F (2,380) - 13.33, p_ <5001) Pairwise comparisons between the peripheral means, using the t^ s t a t i s t i c recommended by Games (1971), established that the mean for the negative reciprocal condition was significantly lower than those for the non-reciprocal and positive reciprocal treatment groups (p_<.01). There was no significant difference between experimental conditions 110 regarding the "percentage of common terms" in the characterizations. However, the peripheral components of the individuals' personal stereotypes contained a significantly greater proportion of "social stereotype t r a i t s " (that i s , were less Idiosyncratic i n content) than the core components (Table 10). The similarities between the social stereotype profiles indicate that many of the traits in the individuals' characterizations of the Orkney Islanders were selected on the basis of information that was common to a l l three versions of the stimulus materials. Indeed, the somewhat pastoral image projected by the tr a i t sets suggests that the respondents' perception of the target as a rural group may have been an important determinant of much of the contents of the personal stereotypes. Yet, although the reciprocal stereotype information did not appear to have a predominant influence on the types of traits that were included in the individuals' characterizations, i t did seem to affect the interpersonal agreement regarding the choice of particular items, since there was less uniformity between the negative reciprocal group members than between the respondents i n the positive and non-reciprocal groups regarding their selection of check l i s t descriptors of the Orkney Islanders. Structural attributes - Differentiation and favourableness - Although the contents of the individuals' characterizations did not alter markedly from one experimental group to another, there were significant differences between treatment conditions with respect to the overall favourableness and differentiation of the personal stereotypes (Table 11). I l l An examination of the mean favourableness and differentiation scores in Table 12 suggests that the characterizations of the positive reciprocal respondents were more differentiated and more favourable than those of the negative and non-reciprocal group members? It suggests, too, that there was l i t t l e difference between the differentiation of the personal stereotypes of the non-reciprocal and those of the negative reciprocal 20 individuals, although the former characterizations appear to have been somewhat more favourable than the latter. The analyses of the differentiation and favourableness scores also indicate that the peripheral components of the personal stereotypes included a significantly greater number of words and were significantly more favourable than the core components (Table 11 and Tableil2). Thesee findings, together with those obtained from the analyses of the common-ness scores, replicate and addeto the results of earlier measurement-comparison studies (e.g., Ehrlich and Rinehart, 1965; McTiernan, 1973). In so doing they add to the empirical support for the conceptual  distinction between the different personal stereotype components (cf. Chapter Two). Further information about the differentiation of the personal stereotypes i s provided by an analysis of the number of favourable, neutral, and unfavourable words in the respondents' characterizations. The means are presented In Table 13 and the analysis of variance is summarized in Table 14. The analysis yielded a significant three-way interaction which is represented graphically in Figure 2. The interpretation of this interaction 112 Table 11. Analyses of variance of the differentiation and favourableness scores Differentiation Favourableness Source MS F MS F Between Ss Experimental group (A) 2 196.93 * 5.08 29.27 ** 10.74 Error 190 38.79 2.72 Within Ss_ Stereotype component (B) 1 10659.89 ** 326.26 19.75 ** 16.64 A X B 2 35.04 1.07 0.28 0.24 Error 190 32.67 1.19 ^ i t h only two levels of the within-subjects variable the degrees of freedom for the conservative F_ tests are the same as those for the conventional tests (cf. Winer, 1971, p. 526) *£ = .007 £ <.0005 113 Table 12. Mean differentiation and favourableness scores for the core and peripheral components of the personal stereotypes of the Orkney Islanders Variable Stereotype Component Positive reciprocal . group (n=58) Non-reciprocal group (n=67) Negative reciprocal group (n-68) Differentiation Core Periphery 6.4 18.1 5.2 15.1 5.0 14.9 Core Favourableness 6.5 6.2 5.5 Periphery 7.0 6.5 6.0 114 was facilitated by tests of the simple simple main effects and, in appropriate instances, by pairwise multiple comparisons using Games' (1971) 21 t s t a t i s t i c . —o Differences between groups There were no significant differences between treatment conditions with respect to the number of favourable (F (2,556) = 2.13, ns), neutral (F (2,556)<1), and unfavourable (F_ (2,556)<1) words in the core components of the respondents' stereotypes: nor were there significant differences with respect to the number of neutral (F (2,556) - 3.64, £ ^ . 0 1 2 2 ) , and negative (F (2,556) - 2.62, ns) words in the peripheral components. However, the respondents i n the positive reciprocal condition did include substantially more favourable words in the peripheral elements of their characterizations than did the members of the other experimental groups (F (2,556) = 29.69, £<£.0005- £<.01 for the differences between the positive and negative reciprocal, and the positive and non-reciprocal means, respectively). Differences between stereotype components The members of a l l three experimental groups included more favourable and neutral words in the :~ peripheral components of their personal stereotypes than they did i n 23 the core components. Moreover, the individuals i n the negative reciprocal treatment condition, in contrast to the respondents in the other groups, also tended to include a greater number of unfavourable items i n the periphery than they did in the core of their characterizations«(F (1,190) = 5.82, £<.05). It should be noted, however, that this tendency was s t a t i s t i c a l l y weak, and should be interpreted with caution given the large number of F_ values being reported. F a v o u r a b l e N e u t r a l U n f a v o u r a b l e F a v o u r a b l e N e u t r a l U n f a v o u r a b l e FAVOURABLENESS CATEGORY FAVOURABLENESS CATEGORY CORE PERIPHERY F i g u r e 2. Mean number o f f a v o u r a b l e , n e u t r a l , and u n f a v o u r a b l e words i n the c o r e and p e r i p h e r y o f t h e p e r s o n a l s t e r e o t y p e s o f t h e members o f each o f the t h r e e e x p e r i m e n t a l groups 116 Differences within stereotype components There was a marked consistency across treatment conditions regarding the general composition of the personal stereotypes. Pairwise comparisons (performed, in turn, on each of the three sets of experimental group means) Indicated that there were more favourable than neutral ( a l l p_s<.05), more favourable than unfavourable ( a l l p_s<.01), and more neutral than negative words ( a l l ps<.01) in the peripheral elements of the respondents' characterizations. There were also significantly more positive than neutral (p_<.05), and significantly more positive than negative (p_<.01) words in the core components of the positive reciprocal group members' descriptions. However, there were no s t a t i s t i c a l l y significant differences between the mean numbers of favourable, neutral, and unfavourable Items in the core of the negative and non-reciprocal respondents' personal stereotypes. A number of points emerge from these findings. F i r s t , there were more pronounced differences between groups with respect to the differentiation of the peripheral stereotype components than there were regarding the differentiation of the core components. Indeed, the greater overall differentiation of the positive reciprocal group members' characterizations was due largely to the individuals' inclusion of a relatively large number of favourable terms in the peripheries of their personal stereotypes (Figure 2). The highly differentiated characterizations of the positive reciprocal group members are related not only to the inclusion of socially desirable information i n the documentary script, but also to the respondents' own post-film judgements that the Orkney Islanders are economically developed and share the same l i f e s t y l e as Canadians. Thus, as Rokeach (1960) has suggested, and as data reported by Smith, Bruner, and White (1956) seem Table 13. Mean number and percentage of positive, neutral, and negative words in the core and peripheral components of the group members' characterizations of the Orkney Islanders CORE PERIPHERY *" "* ——•———---————^———————————— GROUP Variable Treatment Favourableness category Favourableness category MEANS condition p o s^ t_ Neut- Negat- Component Posit- Neut- Negat- Component ive r a l ive means Ive r a l ive means Positive reciprocal 3.40 1.91 1.05 2.12 11.71 4.78 1.66 6.05 4.09 gp. (n=58) Non-reciprocal 2.58 1.60 1.05 1.74 7.72 6.03 1.36 5.04 3.39 Number gp. (n=67) of words Negative reciprocal 2.19 1.52 1.27 1.66 7.74 4.53 2.66 4.98 3.32 gpv,(n=68) Total sample 2.69 1.66 1.12 1.82 8.92 5.12 1.91 5.32 3.57 (N-193) Table 13 (continued) CORE PERIPHERY GROUP Variable Treatment Favourableness category Favourableness category MEANS condition p o s ^ t _ vvu^ut- Negat- Component Posit- Neut- Negat- Component ive r a l ive means ive r a l ive means Positive reciprocal 61.6% 24.5% 13.8% 33.3% 66.2% 25.2% 8.6% 33.3% 33.3% gp. (n=58) Non-reciprocal 52.7% 30.0% 17.3% 33.3% 53.3% 37.7% 9.0% 33.3% 33.3% Percentage gp. (n=67) of words Negative reciprocal 42.6% 29.9% 27.5% 33.3% 50.7% 30.6% 18.7% 33.3% 33.3% gp. (n=68) Total sample 51.8% 28.3% 19.9% 33.3% 56.2% 31.4% 12.3% 33.3% 33.3% (N=193) 119 Table 14. Analyses of variance of the numbers and percentages of favourable, neutral, and unfavourable words in the respondents' characterizations Source df* Numbers of fav., neut., and unfav., words Percentages of fav., neut., and unfav. words MS F b MS F Between Ss Experimental group (A) 2 65.65 5.08 0.05 1.00 Error 190 12.93 0.05 Within Ss Stereotype **** component (B) 1 3553.26 326.25 0.05 1.00 A X B 2 11.68 1.07 0.04 0.80 B X Subj within gps. 190 10.89 0.05 Favourableness **** **** category (C) 2 1855.27 102.49 147550.56 A X C 4 115.69 6.39 8617.73 8.05 C X Subj within gps. 380 18.10 1070.52 **** ** B X C 2 741.91 6 9' 6 5** 4032.59 7.37 A X B X C 4 57.08 5.36 551.84 1.01 B X C X Subj within gps. 380 10.65 546.93 The degrees of freedom list e d are those for the conventional J_ tests. The modified degrees of freedom for the Conservative F tests of the B, C, and B X C effects are 1 and 190, while the conservative degrees of freedom for the between-within interaction effects are 2 and 190 (cf. Winer, 1971, p. 542) ^The p values for the conservative Ftests are denoted p •'-cons *£ = .007 ** *** **** p <.01 .-'-cons p <.001 .-'-cons ^  <.0005 :ons 120 to indicate, there may be a general relationship between the degree to which concepts of social targets are differentiated and the extent to which the targets are perceived to be similar to the individual's most salient reference group or reference person. Although they varied with respect to their overall differentiation, the personal stereotypes were relatively uniform in composition across a l l three experimental conditions. In general they included more positive than neutral, and more neutral than negative words. This finding corres-ponds well with the research that has demonstrated a positive bias in individuals' representations of social stimuli and social relationships (e.g., Morrissette, 1958; Zajonc and Burnstein, 1965; Johnson and Downing, 1976). However, the prevalence of the positive bias i n the responses of the negative reciprocal group members is somewhat surprising in view of the respondents' awareness of the unfavourable information in the film script. The scores representing the numbers of favourable, neutral, and unfavourable words i n the core and periphery of the personal stereotypes were transformed into percentages and reanalyzed. The mean percentage scores are presented in Table 13 and the analysis of variance i s summarized in Table 14. There were two significant interactions. The one between the Experimental group and Favourableness category variables is plotted in Figure 3, and the one between the Stereotype component and Favourableness category variables i s plotted in Figure 4. The post hoc analyses of the interactions provide further information about the structural characteristics of the personal stereotypes, particularly with reference to the previously noted difference between 121 the favourableness of the core and peripheral components. The analyses also provided a clearer understanding of the effect of the reciprocal stereotype stimulus materials on the respondents' tendency to favourably characterize the Orkney Islanders. The peripheral components of the personal stereotypes had a significantly highernpercentage of favourable terms (F (1,190) = 5.06, p_^.05) and a significantly lower percentage of unfavourable terms (F (1,190) = 14.66, jp_^.0005) than the core components (see Figure 4). These findings account for the overall difference in favourableness between the core and periphery of the respondents' characterizations (see Table 11). There was no significant difference between the percentage of neutral words in each of the stereotype components (F (1,190) = 2.44, ns). Although the core and periphery of the personal stereotypes differed in composition they were structurally similar to one another (see Figure 4). That i s , they both included a greater percentage of favourable than neutral (p_<.01, in both cases), and a greater percentage of neutral than negative words (p_.< .01, i n both cases) . Indeed, with one qualification, this general finding held for the characterizations provided by the members of each of the three experimental groups. The exception occurred in the stereotypes of the individuals in the negative reciprocal treatment condition. Unlike the other respondents, these cadets did not provide a significantly greater proportion of neutrall than negative terms when describing the target group (see Figure 3). However, while the broad structural features of the characterizations did not vary a great deal across experimental groups, there were substantial differences between treatment conditions with respect to the Favourable Neutral Unfavourable FAVOURABLENESS CATEGORIES Figure 3. Mean percentages of favourable, neutral, and unfavourable terras in the personal stereotypes of the positive, negative, and non-reciprocal group members 100 Figure 4. Mean percentages of favourable, neutral, and unfavourable items in the core and peripheral components of the respondents' personal stereotypes 124 exact proportions of favourable, neutral, and unfavourable items in the respondents' stereotypes. The members of the positive reciprocal group included a higher percentage of favourable words In their characterizations than did the individuals in the non-reciprocal (p_^.05) and negative reciprocal (p_< .01) conditions (F (2,380) = 13.36, p_<.0005). Likewise the respondents in the negative reciprocal condition incorporated a higher percentage of unfavourable terms into their personal stereotypes than did the members of the positive (p_< .01) and non-reciprocal (p_<^ .05) groups (F "(2,380) = 7.30, p_<.001). There was also a marginally significant difference between groups with respect to the percentage of neutral words in the individuals' responses (F (2,380) = 3.65, p_<.05). In this instance the highest mean (33.9%) was associated with the non-reciprocal group and the lowest mean (24.9%) with the positive reciprocal condition (see Figure 3). These findings reflect the extent to which the positive reciprocal stimulus materials enhanced the general tendency of the respondents to ascribe favourable traits to the Orkney Islanders. Indeed, favourable ascriptions account for an average of sixty-four percent of the descriptors provided by the individuals in the positive reciprocal condition, and this i s not only significantly but substantially more than the proportion of favourable terms in the stereotypes of the non-reciprocal (M = 53.0%) and negative reciprocal (M = 46.7%) group members. It i s hardly surprising to find that the high proportion of favourable descriptors in the responses of the individuals in the positive and 125 non-reciprocal groups was associated with a low. percentage of unfavourable terms. The tendencies to l i s t a large proportion of positive items and to provide relatively few unfavourable terms when describing a social target appear to be two components of a more general modus operandi whereby individuals describe others in as desirable a manner as possible in a l l situations except those i n which a tolerance of negative commentary i s well-defined or can readily be assumed. The negative reciprocal feedback served to lessen individuals' adherence to the norm against providing a large number of unfavourable terms when characterizing social agents. And while the positive bias was not f u l l y eliminated by the relatively high percentage of negative terms, the inclusion of a substantial portion of unfavourable items in the characterizations provided by the negative reciprocal respondents did lower the average favourableness ratings of these stereotypes. This effect i s particularly noticeable when the mean scores for the negative reciprocal condition are compared with those for the positive reciprocal group (Table 12). As was mentioned previously, the presence of a weakened but s t a t i s t i c a l l y detectable positive bias in the responses of the negative reciprocal group members i s an unexpected result. It diverges sharply from findings i n the impression formation literature which have demonstrated the disproportionate influence of negative information on evaluations of target persons (e.g., Birnbaum, 1972; Hamilton and Huffman, 1971; Riskey and Birnbaum, 1974). For example, Richey, Koenigs, Richey, and Fortin (1975) demonstrated that 126 where the amount of positive information was greater than the negative, results indicated that a single negative behaviour neutralized five positive behaviours, yielding an impression only Insignificantly better than that based on five negative and five positive actions (p. 233), while Hamilton and Zanna (1972) found that likableness ratings of the stimulus person described by an unfavorable attribute were significantly more discrepant from a neutral Impression than were ratings of the stimulus person characterized by a desirable attribute (p. 204). One possibility that must be considered when attempting to explain the difference between the results of this study and the findings of those in the Richey and Hamilton tradition (e.g., Gray-Little, 1973; Cusumano and Richey, 1970; Richey and Dwyer, 1970; Hamilton and Huffman, 1971; Hamilton and Zanna, 1972) is that theypresent stimulus materials were structured i n such a fashion that they minimized the impact of the negative reciprocal information. The fact that a l l of the reciprocal stereotype feedback was included towards the end of the documentary scripts suggests that a primacy effect (e.g., Anderson, 1965) may have been operating when the negative reciprocal group members formed their stereotypes of the Orkney Islanders. Faced with two apparentlyy incongruent pieces of information about the Islanders - that they are a mild-mannered people who conduct their business without fuss, and that they passionately dislike Canadians -the respondents may have attempted to reduce the incongruenceLby discounting the reciprocal stereotype feedback and dwelling almost exclusively on the more general documentary information at the beginning of the film. There is no evidence of such a discounting process,however. 127 The manipulation checks indicate that not only did a majority of the individuals in the negative reciprocal condition clearly express their awareness of the Orcadians' "views of Canadians," a number also used this information as a focus for their comments in the post-film questionnaire,. For example, one individual ventured that the Islanders see other people in a manner "similar to how many Canadians view Americans: they blame us for past shortcomings . . . . I don't like people who don't like me." A primacy effect explanation of the present findings is also countered by the results of Richey's research. She has found repeatedly (e.g., Richey, McClelland, and Shimkunas, 1967) that when the experimental procedure is designed to encourage respondents to attend equally to a l l sections of the stimulus materials, as i t was in this study (see Chapter Four), a sequence of favourable followed by unfavourable phrases produces definite negative impressions which persist over a considerable period of time (7 to 9 days, at least). Thus, i t i s d i f f i c u l t to argue that the structure of the stimulus materials was singularly responsible for the differences between our findings and those of the comparison studies. Rather we must look for other features of these materials which may account for the discrepancies. The present stimulus materials can be distinguished from those of the "traditional" impression formation studies in terms of the quantity of information provided to the respondents, the realism of the target, and the use of attitudinal rather than behavioural or dispositional information as the independent variable. 128 The relatively large amount of information presented in this study involved the participants being exposed to the target stimulus for a much longer period of time than is normally the case. It also entailed their being provided with a context within which to react to the independent variable. It is possible, then, that the uncharacteristically weak impact of the negative stimulus materials may have been due to the effects of mere exposure to the target (Zajonc, 1968), or to the respondents having had the opportunity to evaluate the reciprocal stereotype feedback with reference to other related material rather than in an informational vacuum. The weak impact of the negative reciprocal feedback may also have been due to the fact that the target was real rather than f i c t i t i o u s . In most studies of person perception and impression formation, whether they involve the presentation of t r a i t l i s t s , short descriptive paragraphs, or brief audio-visual documentaries, the target persons (or, less frequently, groups) are defined, albeit implicitly, as f i c t i o n a l characters, much like those in novels and movies. Strong reactions to such targets that have been negatively described may well be determined by this assumption of their unreal status. The risks involved i n harshly judging a f i c t i o n a l character are much less than those involved in overtly c r i t i c i z i n g an acquaintance in a social context that may not be supportive of such a criticism. Indeed, as has been suggested already, i t perhaps the case that certain norms for tolerating criticism must be perceived to be present before most people w i l l openly venture to derogate individuals with whom they feel uncomfortable. When such norms are not evident people w i l l tend to employ more guarded responses (e.g., Sigall \ 129 and Page, 1971).. Finally, the most apparent explanation of why the negative reciprocal materials did not have a disproportionately strong effect on the contents of the personal stereotypes i s that the respondents appraised the unfavourable attitudinal information differently than they would have evaluated similar behavioural or dispositional information. It i s conceivable, for example, that the target's attitudes towards Canadians were considered to be potentially modifiable and were consequently judged to have less serious social implications than acts or traits that imply long term and stable dispositions (. see Jones and Davis (1965), Kanouse and Hanson (1972), Marston (1976), and Wiggins (1974) for related commentaries). It may also be the case that individuals' verbal association networks are so structured that the task of attributing a variety of unfavourable traits oh the basis of specific, negative attitudinal information was a more d i f f i c u l t one than the task of making such attributions on the basis of specific dispositional information. Tests of each of these possible explanations w i l l help to define the parameters of the stereotype formation model under discussion. In this respect i t is interesting to note that in a study in which respondents were provided with details of a positive or a negative act performed by an individual with whom they had interacted for five minutes, Weinstein and Crowdus (1968) failed to support their hypothesis that negative information has more saliency than positive information. 130 It i s also worth-noting Marston's (1976) finding "that i n making-statements of t r a i t attribution to persons the observer i s capable of assimilating and integrating a considerable amount of Information (p. 254)7" Her conclusion is particularly relevant: The present findings do suggest the need for more careful empirical study of the processes involved In making person attributions before investigators are just i f i e d i n concluding that observers typically use a very limited amount of information i n making dispositional attributions. More specifically, i t points to the need for experimental designs which expose the observer to a considerable amount of information, rather than the usual designs involving brief exposure to a single action about which the observer is given only minimal information, (p. 255) Respondents' Use Of stimulus material information The discussion of the favourableness, differentiation, and commonness scores provides information about the evaluative characteristics of the traits attributed to the Orkney Islanders, and about the manner in which the traits were structured within the respondents' personal stereotypes. It remains to examine the extent to which these stereotype traits were borrowed by the respondents from the stimulus descriptions of the target group. An index of social learning was derived by counting the overlap between the words in the core component of each individual's characterization of the Islanders and the twenty traits that were included in the film script representation of the target. A similar count was performed on the items in the peripheral components of the respondents' stereotypes. An analysis of the scores does not provide any evidence that the individuals tended to include stimulus material descriptors of the target 131 in the core components of their personal stereotypes. The mean number of inclusions was between 0.2 and 0.3 for each of the treatment groups. Nor do the findings indicate that the respondents recognized many of the neutral film descriptors as being particularly characteristic of the Orkney Islanders when they were presented with these twenty tr a i t s on the check l i s t . The individuals In the positive reciprocal condition attributed an average of 3.8 of the terms to the Islanders when completing the check l i s t task. The non-reciprocal group members checked an average of 3.4 of the items, and the mean for the negative reciprocal condition was 3.9. In brief, i t appears that the content of the respondents' personal stereotypes was not markedly influenced by memory or recognition of those traits that were used to characterize the target in the stimulus films. A related issue emerges with respect to the reciprocal feedback treatment conditions. While the individuals in these groups did not use many of the film descriptors of the target in their own characterizations of the Orkney Islanders they may have included a number of the labels that were supposedly attributed to Canadians by the Orcadians. However, the scores denoting the overlap between the contents of the respondents' characterizations and the contents of the favourable and unfavourable film representations of Canadians indicate that this was not the case. The core components of the positive reciprocal group members' stereotypes contained few of the complimentary film descriptors of Canadians (M = 0.6). Likewise, the undesirable terms that had been attributed to Canadians in the unfavourable feedback condition were seldom included in the core of the negative reciprocal group members' character-izations (MM= 0.1). 132 Overlap indexes for the check l i s t data were not analyzed since an interpretation of the results would have been inconclusive. Overlap between the content of the respondents' stereotypes of the Orkney Islanders and their concepts of Canadians Finally, the relationship of the respondents' newly formed stereotypes of the Orkney Islanders to other concepts in their personal stereotype systems was examined by measuring the content overlap between the characterizations of the Orcadians and those of Canadians, Germans, and the Arabs.- the three pseudo-targets in the study. Although there was some overlap between the terms attributed to the Orkney Islanders and the terms ascribed to the other national-ethnic groups, the contents of the Orkney Islander stereotypes were distinctive and substantially different from the contents of the Canadian, German, and Arab representations. Regarding the content overlap, however, there were more traits in common between the positive reciprocal respondents' characterizations of the Orcadians and Canadians (M = 11.3) than between their characterizations of the Orcadians and Germans (M = 10.0). Likewise, there were more terms in common between their stereotypes of the Orkney Islanders and the Germans than between their stereotypes of the Orkney Islanders and the Arabs (M = 6.3). This general pattern of findings also held for the characterizations of the non-reciprocal group members, although the mean content overlap scores were somewhat lower than those for the positive reciprocal condition. For example, while the mean number of traits i n common between 133 the Orkney Islander and German stereotypes was 10.0 for the members of the positive reciprocal group, i t was 6.6 for the non-reciprocal treatment condition. Furthermore, the members of the non-reciprocal group had more terms InuCOimTionibetween their Canadian and Orcadian stereotypes (M = 8.4) than did the respondents In the negative reciprocal condition (M = 6.8). The negative reciprocal individuals, on the otherhand, provided more overlap between their Arab and Orkney Islander characterizations (M = 5.3 for the negative reciprocal condition, M = 4.0 for the non-reciprocal group). The mean scores for the number of items in common with the German stereotypes were approximately the same for both the non-reciprocal (M = 6.6) and negative reciprocal (M = 6.8) treatment groups. Proportionately more of the content overlap occurred between the peripheral components of the personal stereotypes than between the core components. This i s an intuitively plausible finding since the traits which form the core of individuals' stereotypes can be expected to be those which mark the most distinctive features of the target in question. In essence, these findings suggest that there were differences between treatment groups with respect to the general relationship between the respondents' characterizations of the Orkney Islanders and the other concepts in their personal stereotype systems. More particularly, there was much less content similarity between the negative reciprocal group members' concepts of the Orkney Islanders and Canadians (M = 6.8) than there was between the positive reciprocal respondents' characterizations of these target groups (M = 11.3). The 134 mean overlap score for the non-reciprocal condition was 8.4. An analysis of the content overlap between the favourable t r a i t s in the individuals' stereotypes clearly illustrates the variation between the overall scores obtained from the different experimental groups. The mean number of favourable terms in common between the positive reciprocal group members' characterizations of the Orkney Islanders and Canadians (M =8.8) was significantly higher than the means for the non-reciprocal (M = 5.5) and negative reciprocal (M = 4.6) groups. These differences between means were significant at the .01 level. The F_ for the simple simple main effect (Experimental group X Target X Favourableness category was the significant interaction on which the simple simple main effect was calculated) was 25.78, df = 2,671, p_<.0005. One of the broad implications of Campbell's (1967) wide ranging discussion of stereotype systems i s that the greater the real or apparent contrast between target groups on any salient social dimension the less similar w i l l be the contents of the individual's concepts of these groups. The present results support this speculative relationship in that the content overlap between the respondents' stereotypes of the Orkney Islanders and Canadians was greater for those who were led to believe that the Islanders view Canadians very favourably than i t was for those^ who were not given any information concerning the Orcadians' attitude towards Canadians, or for those who were told that the Islanders strongly dislike Canadians. 135 Review of the analyses of the personal stereotype variables The analyses of the dependent variables indicate that there was considerable similarity in the types of words that were attributed to the Orkney Islanders by the members of the different treatment groups. Indeed, much of the content of the respondents' newly formed stereotypes appears to have been determined by the individuals' perceptions of the target as being a rural group. The content similarity of the characterizations e l i c i t e d from the individuals in the different experimental conditions was reflected in the general structure of their stereotypes. By and large the individuals i n c l included a greater proportion of favourable than neutral words, and a greater proportion of neutral than negative terms in their descriptions of the Orkney Islanders. However, while the personal stereotypes possessed the same overall structural features regardless of treatment condition, their form varied as a function of the particular feedback about the Orkney Islanders' stereotypes of Canadians. The personal stereotypes of the positive reciprocal group members were more differentiated and more favourable than those of the negative reciprocal respondents. They included a greater number and a greater percentage of favourable items, and a smaller number and a smaller percentage of unfavourable terms. Moreover, there was also more agreement among the individuals in the positive reciprocal condition than there was among the members of the negative reciprocal group concerning which particular traits could be considered to be characteristic of the Orkney Islanders. The stereotypes of the non-136 reciprocal respondents f e l l between those of the positive and negative reciprocal group members on a l l dimensions. Generally speaking, while the differences between groups were apparent in both components of the personal stereotypes, they were more pronounced in the periphery than in the core. c. Canonical correlation analyses of the relationship between the personal  stereotype variables and the estimated reciprocal stereotype measures The results outlined in the previous sections have established that the respondents made inferences about the Orcadians' stereotypes of Canadians in the absence of definite information about the nature of these reciprocal stereotypes; that they attended to such reciprocal stereotype information when i t was included i n a documentary presentation about the Islanders; and that the form of the respondents' estimates of the Orcadians' stereotypes of Canadians, as well as the form of their own characterizations of the Islanders varied as a function of the type of information presented to them. One additional and fi n a l set of analyses was performed to extend these findings and to establish the degree of Interrelatedness between the respondents' scores on the personal stereotype and estimated reciprocal stereotype variables. Using canonical correlation analyses i t was possible to identify the set of linear combinations of personal stereotype variables that correlated most highly with linear combinations of the estimated reciprocal stereotype variables. As Cooley and Lohnes (1971) indicated: 137 The canonical correlation model uses the same analytic trick to display the structure of relationships across domains that the factor model uses to display the structure of relationships within a domain The canonical model selects linear functions that have maximum covariances between domains, subject to restrictions of orthogonality, (p. 169) Two canonical correlation analyses were conducted for each of the three treatment groups. The different analyses for each group included slightly modified subsets of personal stereotype and estimated reciprocal stereotype variables. The complete set of personal and estimated reciprocal stereotype measures contain variables that are linearly dependent. For example, given the differentiation scores and the numbers of favourable and neutral words in the respondents' stereotypes i t i s possible to calculate the number of negative terms in their characterizations. Thus the latter variable is a linear function of the former measures. However, linear dependence within a data matrix Is a condition which renders insoluble a number of matrix algebra operations, including those essential to the solution of canonical correlation equations. In order to overcome this problem the measures of the number and percentage of unfavourable terms in the core and periphery of the characterizations were excluded from the f i r s t set of analyses. They were replaced in the second set while the measures of the number and percentage of neutral terms were deleted. The outcomes of the canonical analyses were consistent from one experimental condition to another and from one subset of variables to the other. Each of the six analyses produced three significant canonical correlations. A scrutiny Of the standardized canonical variate coefficients 138 associated with, these significant correlations suggests that in general the more differentiated and favourable were the positive, negative, and non-reciprocal group members' estimates of the Orkney Islanders' stereotypes of Canadians the more differentiated and favourable were their own stereotypes of the Orkney Islanders. An example of the findings i s presented in Table 15, which summarizes the f i r s t of the two canonical correlation analyses of the non-reciprocal respondents' scores on the personal and estimated reciprocal stereotype variables. The table includes the significant canonical correlations and the matrices of standardized, normalized canonical variate coefficients. These matrices of coefficients provide exact details about the composition of the three pairs of canonical variates (linear combinations of variables) that produced the significant canonical correlations. The relative magnitude of the coefficients indicates the contribution of each of the personal and estimated reciprocal stereotype variables to their respective canonical variates. The coefficients for the f i r s t two pairs of canonical variates reflect a relationship between the favourableness and differentiation of the personal stereotype peripheries and the favourableness and differentiation of both components of the estimated reciprocal stereotypes of Canadians. The f i r s t pair of canonical variates seem to represent a favourableness-differentiation dimension. Low percentages of favourable words and low levels of differentiation in the peripheral components of the personal stereotypes were associated with low percentages of favourable words in the peripheries and small numbers of words in the core components of the estimated reciprocal stereotypes of Canadians. Similarly, highly favourable Table 15. Canonical correlations of personal stereotype and estimated reciprocal stereotype variables for the non-reciprocal group — negative variables excluded Canonical Wilk's Chi-Eigenvalue correlation Lamda Square df £ 1 0.716 0.846 .0014 339.30 224 <.0005 2 0.674 0.821 .0048 274.40 195 <.0005 3 0.638 0.799 .0149 216.71 168 <.006 Table 15 (continued) Matrices of standardized, nor Personal stereotype variables 1 2 3 Core Differentiation -0.041 0.007 0.052 Favourableness -0.097 0.041 -0.468 % fav. words 0.176 0.002 0.607 % neutral words 0.120 0.037 0.347 No. fav. words -0.091 -0.013 -0.048 No. neutral words -0.015 -0.004 -0.030 Commonness scores -0.006 0.025 -0.083 Periphery Differentiation -0.414 -0.568 -0.178 Favourableness 0.252 0.072 -0.230 % fav. words -0.612 -0.033 0.311 % neutralr-swords -0.357 -0.347 0.218 No. fav. words 0.262 -0.071 0.167 No. neutral words 0.359 0.735 0.149 Commonness scores 0.029 -0.016 -0.026 canonical coefficients Estimated reciprocal stereotype variables 1 2 3 Core Dif ferentiation -0.452 0.480 -0.190 Favourableness 0.227 XK178 -0.413 % fav. words -0.285 0.046 0.506 % neutral words -0.184 0.138 0.140 No. fav. words 0.217 -0.361 0.109 No. neutral words 0.198 -0.197 0.214 Commonness scores 0.103- -0.022 2=0.059 % common terms -0.133 0.090 -0.030 Periphery Dif ferentiation 0.145 -0.389 -0.099 Favourableness 0.328 -0.120 -0.182 % fav. words -0.516 -0.060 0.036 % neutral words -0.197 -0.275 -0.028 No. fav. words 0.029 0.125 0.413 No.aneutral words -0.029 0.496 0.312 Commonness scores -0.248 -0.146 -0.343 % common terms 0.136 0.106 0.157 141 and highly differentiated (i.e., high, loadings on the "number of favourable and neutral terms" variables) personal stereotype peripheries were associated with very favourable peripheral components of the estimated reciprocal stereotypes. The coefficients for the second pair of canonical variates indicate -that when the individuals' estimated reciprocal stereotypes included differentiated core components and a large number of neutral words in the peripheral segments, their personal stereotype peripheries also included a substantial number of neutral words. Likewise, estimated reciprocal stereotypes that had relatively undifferentiated peripheries and had few positive terms in their core components were associated with personal stereotypes whose peripheral components included few words and low percentages of neutral items. This pattern of associations suggests that a differentiation dimension underlies the canonical variates for the second correlation. The third set of canonical variate coefficients displays a strong relationship between the favourableness of the core components of the personal stereotypes and the favourableness of the core components of the respondents' estimates of the Orkney Islanders' reciprocal stereotypes of Canadians. A high proportion of favourable terms in the personal stereotype cores was associated with a high proportion of favourable terms in the core components of the estimated reciprocal stereotypes, while low favourableness scores in the core components of the personal stereotypes tended to go with low favourableness scores in the cores of the estimated reciprocal stereotypes. 142 It i s evident from this example, (and from the other canonical analyses) that the correlation findings are similar to and indeed complement the results obtained from the analyses of variance. d. Review of the findings When taken together both the analysis of variance and the canonical correlation findings indicate that the characteristics of the core and peripheral components of the personal stereotypes varied (1). as a function of the type of reciprocal stereotype information presented i n the stimulus materials, and (2) In relation to the manner in which this reciprocal information was construed by the respondents. Thus the results of the study offer support for the general proposition stated in the introduction (Chapter Three): An individual's inference about a target's characterization of his/her most salient reference group acts as a mediating variable i n the stereotype formation process. Chapter overview The results of the manipulation checks were presented in the f i r s t section of this chapter as were the findings of a number of preliminary methodological analyses including a "data-pooling" pretest, homogeneity of variance tests, and the correlations between age and thespersonal stereotype variables. The respondents' post-film impressions of the Orkney Islanders and their estimates of the Islanders' reciprocal stereotypes of Canadians were discussed in later sections of the chapter, as were the analyses of the contents and structure of the respondents' personal stereotypes of the Orkney Islanders, and the canonical correlation analyses of the relationship between the personal stereotype variables and the estimated reciprocal stereotype measures. 143 CHAPTER SIX: SUMMARY AND CONCLUSIONS Research- findings suggest that the contents and structure of individuals' stereotypes are determined largely by reference group norms. These norms can include open-ended t r a i t association and target Identification rules in addition to cultural definitions of particular, socially salient targets. It appears that, when necessary, individuals order the association and identification rules into stereotype formation strategies that w i l l f a c i l i t a t e the characterization of unfamiliar and socially non-attended targets. Thus, a comprehensive account of the stereotype formation process requires at least two models, one defining the variables which govern the development of stereotypes about targets possessing clear cultural definitions, and the other delineating the form of individuals' more general "stereotype formation strategies." The present study was concerned with the presentation of a preliminary, three-step model of the general stereotype formation process and with the results of an experiment which was designed to test some of the assumptions underlying the model. The model postulates that: individuals use available information to construct demographic profiles of culturally undefined targets; attribute particular "world views" and value systems to the targets on the basis of the inferred demographic characteristics; and then make t r a i t ascriptions on the basis of their inferences about the targets' general perspectives, using reference group norms as their judgemental anchors in the process. 144 Person perception studies indicate that individuals' perceptions of how a target characterizes their own reference group may be one such "world view" variable which mediates i n the stereotype formation process. The present study was designed to explore this possibility by providing answers to the following questions: Will respondents make inferences about an unfamiliar target group's (the Orkney Islanders)sstereotypes of their own national group in the absence of definite Information about the nature of these "reciprocal stereotypes"? Will such information be attended to when i t i s included i n a documentary script about the Orkney Islanders, and w i l l i t affect the form and structure of the core and peripheral components of the newly developed stereotypes? Three experimental conditions were established by pairing a film about l i f e on the Orkney Islands with each of three versions of a tape recorded, documentary passage. The basic passage contained an account of the history, economic a c t i v i t i e s , and l i f e s t y l e of the people, In addition to a description of the Islanders which was built around twenty preselected, neutral t r a i t words (non-reciprocal condition). The other two variations of the documentary script included information concerning the manner in which the Orkney Islanders "characterize Canadians." This information replaced non-essential padding material from the original script. The Islanders were described as viewing Canadians very favourably in the positive reciprocal stereotype condition, and twenty positive attributes were used as examples of traits which were frequently assigned to Canadians. Alternatively, in the negative reciprocal stereotype condition the Orkney Islanders were described as viewing Canadians very unfavourably. The 145 negative antonyms of the positive traits were used in this instance. The film sequences lasted for twenty-four minutes each - the length of a short TV programme. One hundred and ninety-three National Sea Cadets from H.M.C.S. Quadra at Comox, B.C., completed a l l stages of the experiment. They f e l l naturally into nine groups which were allocated randomly to one of three experimental conditions. Fifty-eight of the participants saw the positive reciprocal film, sixty-seven saw the control film, and sixty-eight were shown the negative reciprocal version of the presentation. Each of the three groups in each of the experimental conditions saw the relevant film under the guise of participating in an education research project. The task, as i t was explained to the respondents, was to evaluate and suggest improvements on a preliminary version of a film that was being prepared for use in introductory social science courses. The individuals were asked to watch the film and li s t e n to the taped soundtrack carefully, and then to assess the merits and shortcomings of the presentation on a questionnaire that was distributed at the end of the viewing period. The questionnaire included items relating to the respondents' f i r s t impressions of the target group. One week after they had been exposed to the stimulus materials and had completed the post-film questionnaire the individuals were recalled to the " f i e l d laboratory" and asked to complete a four-part stereotyping task using a free response as well as a check l i s t instrument. The participants were required i n i t i a l l y to characterize Canadians, Orkney Islanders, Germans, and Arabs i n their own words and then to repeat the task while playing the role of an Orkney Islander. In addition 146 they were asked to rate each of the terms that they had lis t e d on a nine-point favourableness scale. Finally, the entire free response sequence was repeated with the adjectival check l i s t format. The order of presentation of the targets was randomized within each of the four sections of the task. The check l i s t contained the twenty positive, neutral, and negative t r a i t words that were included i n the stimulus materials, as well as fifteen other words that were selected from the Karlins, Coffman, and Walters (1969) adjective l i s t . The experimentally formed stereotypes and the respondents' estimates of the Orkney Islanders' reciprocal stereotypes of Canadians were examined in terms of their favourableness, their differentiation, their inclusion of commonly used words, and their inclusion of descriptive t r a i t s that were contained in the stimulus materials. Chi-square analyses were employed to explore the relationships between experimental group membership and responses on the post-film questionnaire. Univariate, unweighted means, between-within analyses of variance were performed on the personal stereotype variables. Experimental groups and stereotype components (core measured by the free response and periphery measured by the check l i s t format) were the between- and within-subjects independent variables, respectively. Canonical correlation analyses were employed to examine the relationships between the personal stereotype variables and the measures obtained from the respondents' estimates of the reciprocal characterizations of Canadians. The canonical analyses were performed separately for each of the three experimental groups. 147 The manipulation checks and the analyses of the respondents' estimates of the Orkney Islanders' reciprocal stereotypes of Canadians indicate that even i n the absence of definite cues about the nature of these reciprocal stereotypes, the non-reciprocal group members made inferences about the Orcadians' perceptions of Canadians. The analyses also indicate that the respondents' i n the positive and negative reciprocal conditions attended to the reciprocal stereotype information embedded in the stimulus materials. They made correct judgements about the evaluative tone of the information, and they employed many of the tra i t s comprising this reciprocal information when they estimated the Orkney Islanders' characterizations of Canadians. The analyses of the personal stereotype variables further determined that the individuals who received positive reciprocal feedback about the Orkney Islanders' stereotypes of Canadians had personal stereotypes of the target group that were more differentiated and more favourable than those of the respondents who attended to the negative reciprocal stereotype Information. The characterizations of the former group included a greater number and a greater percentage of favourable terms and a smaller number and smaller percentage of unfavourable Items than those of the latter : individuals. Moreover, there was more agreement among the respondents in the positive reciprocal condition than among the members of the negative reciprocal group concerning which particular traits could be considered to be characteristic of the Orkney Islanders. The stereotypes of the non-reciprocal individuals f e l l between those of the positive and negative reciprocal group members on a l l dimensions. 148 While the analysis of variance results established that the form of the respondents' characterizations varied as a function of the type of reciprocal stereotype information presented to them, the canonical correlation results indicate that the form of the personal stereotypes also varied In relation to the respondents' scores on the measures of their estimates of the Orkney Islanders' reciprocal stereotypes of Canadians. In general, the more differentiated and favourable were the positive, negative, and non-reciprocal group members' estimates of the Islanders' stereotypes of Canadians the more differentiated and favourable , were their own stereotypes of the Orkney Islanders. To review, the findings determined that the respondents made inferences about an unfamiliar target group's stereotypes of their own national group in the absence of definite information about the nature of these "recip-rocal stereotypes;" that reciprocal stereotype information was attended to when i t was included in a documentary script about the Orkney Islanders; and that the characteristics of the personal stereotypes varied (1) as a function of the type of reciprocal stereotype information presented in the stimulus materials, and (2) in relation to the manner in which this reciprocal information was construed by the respondents. Thus the results offer support to the general proposition that: An individual's inference about a target's characterization of his most salient reference group acts as a mediating variable i n the stereotype formation process. Moresgenerally, the pattern of results lends support to the assumptions underlying the three-step model of the stereotype formation process in that the respondents were able to provide clearly defined demographic profiles of the Islanders; were able to attribute world view 149 perspectives to the Islanders on the basis of the available demographic information; and were influenced by such-world view information and world view inferences when making t r a i t attributions to the Islanders. However, a number of the findings suggest that the model, as i t was outlined in the introduction to the study, i s overly simple and requires further elaboration. A schematic diagramfofbthe model i s presented in Figure 5. The solid lines represent the processes that were originally assumed to operate, while the broken lines represent the additional inferential processes that were suggested by the results of the experiment. As can be seen from the figure, the three steps hypothesized to be involved in the formation of stereotypes were i n i t i a l l y assumed to be linearly ordered and non-reflexive. Demographic information was considered to provide sufficient input to allow the individual to generate a detailed demographic profile of the target; to associate particular world view beliefs and values with this profile; and subsequently to generate what he/she thought to be appropriate sets of trait-attitude-behaviour attributions on the basis of these inferences about the target's social.perspectives. A further, unstated assumption was that when information about aspects of the target's world view was also available to the individual, he/she would employ i t only when making inferences about the said target's social beliefs and values and not when constructing the i n i t i a l demographic pro f i l e . A similar assumption was made with regard to the individual's use of tr a i t information. DEMOGRAPHIC INFORMATION I N P U T F R O M S O C I A L E N V I R O N M E N T "WORLD VIEW" INFORMATION TRAIT INFORMATION L E X I C O N O F S O C I A L T E R M S I N T H E I N D I V I D U A L ' S C O G N I T I V E D I C T I O N A R Y "7" DEMOGRAPHIC PROFILE ( e . g . , i n f e r e n c e s about age, s e x , r a c e , s o c i o - e c o n o m i c s t a t u s o f t a r g e t T INFERENCES ABOUT TARGET'S WORLD VIEW ( e . g . , t a r g e t i s p o l i t i c a l l y c o n s e r v a t i v e , e c o l o g y minded) PERSONALITY STEREOTYPE ( e . g . , t a r g e t i s r e s e r v e d , h a r d w o r k i n g , a l t r u i s t i c ) r I o F i g u r e 5. S c h e m a t i c d i a g r a m o f t h e t h r e e - s t e p s t e r e o t y p e f o r m a t i o n model 151 Furthermore, i t was postulated that t r a i t attributions were made on the basis of inferences about the target's world view and not on the basis of demographic information such as race, age, or socio-economic status. However, the analyses of the respondents' post-film impressions of the Orkney Islanders indicate that world view information can combine with demographic information to shape the form of individuals' s e l f -generated demographic profiles of a target. In addition, the pastoral overtones reflected in the social stereotypes of a l l three experimental groups suggest that many of the traits ascribed to the Orkney Islanders were attributed on the basis of the respondents' perceptions of the Orcadians as being a relatively poor, rural group rather than on the basis of their inferences about the Orkney Islanders' view of Canadians. Two points emerge from these findings. The f i r s t i s that constructing a demographic profile and making inferences about a target's world view may not always be discrete and sequential processes, but may often be closely interlinked operations. The second i s that t r a i t ascriptions can be generated as readily from some demographic information as they can from inferences about a target's social perspective (see also Stein, Hardyck, and Smith, 1965). However, whether or not demographic material w i l l be attended to and weighed more heavily than world view material depends on the social context in which the stereotypes are being formed. In some instances i t i s appropriate to focus on age, sex, or occupation as salient and relevant characteristics, and in others i t i s more appropriate to attend to beliefs, values, and p o l i t i c a l orientations (e.g., Insko and Robinson, 1967; Triandis and Davis, 1965). 152 Based on the above results, a schematic representation of a revised and more general stereotype formation model would include a reflexive connection between the 'world view information' and 'demographic pro f i l e ' components. It would also Incorporate a direct link between the 'demographic profile' and 'personality stereotype' components. With further work such a model might additionally be expected to incorporate information concerning the types of social situations in which demographic information would be more salient than world view information, and vice  versa. On a broader level, the experimental findings suggest that any discussion of stereotype modification and change should make reference to whether or not such change i s expected in both the core and peripheral components of the characterizations under study. The present results indicate that alterations In one stereotype component are not always coupled with similar alterations in the other. Lastly, the attempt to account for the lack of a very high proportion of unfavourable terms in the stereotypes of the negative reciprocal respondents highlighted the need to further explore the consequences of using real as opposed to unreal target stimuli, a great deal rather than a limited amount of information, and attitudinal as opposed to behavioural or dispositional material when studying the stereotype formation process. 153 PART THREE OVERVIEW AND FINAL COMMENTS 154 CHAPTER SEVEN: OVERVIEW AND'FINAL COMMENTS Three principal arguments were advanced in the present thesis: that a broad definition of the term 'stereotype' is more tenable than the traditional, and more narrowly restrictive definitions; that free response instruments measure different components of personal stereotypes than structured formats; and that a comprehensive analysis of the formation of stereotypes entails a study of the processes underlying individual's general stereotype construction strategies as well as a study of the processes involved in the learning of cultural definitions of salient social targets. A preliminary, three step model of the general stereotype formation process was presented and the results of an experiment designed to test some of the assumptions of the model were discussed. Some f i n a l comments can help put the experimental findings and the arguments of the f i r s t two chapters in perspective. The definition of stereotypes as being social concepts is predicated on the assumption that under normal circumstances the 'contents' of these concepts remain relatively stable across time, and would be largely unchanged i f el i c i t e d under a variety of different test conditions. However, this assumption has yet to be tested empirically and, indeed, much research on the question remains to be done. The distinction between the core and peripheral components of personal 155 stereotypes i s a theoretically useful one. It helps to explain the discrepancies between results obtained from free response and those obtained from check l i s t measuring instruments, and, more generally, i t emphasizes the importance of maintaining a close correspondence between the theoretical and operational definitions of the stereotype construct. The approach adopted in the stereotype formation study was essentially an information processing one and no account was taken of the role of 'interpersonal interaction' variables in the development of .individuals 1 social concepts. The experimental scenario was analogous to a TV or a movie viewing episode and respondents were actively discouraged from exchanging views and opinions about the film. Yet i t is more than apparent that many concepts about groups and individuals are formed as a function of social interaction. Consequently, a comprehensive model of the?>stereotype formation process must incorporate systematic research on the influence of 'group dynamic' variables on the development of stereotypes. A start on such work has been made by researchers such as Avigdor (1953) and Sherif, Harvey, White, Hood, and Sherif (1961). Finally, in broad terms, this thesis i s similar to most other stereotype research i n that i t f a l l s squarely within the associationist tradition of cognitive social psychology. Hopefully, the general orientation toward the f i e l d that has developed out of this tradition can be complemented in the future by alternative perspectives that are founded in other theoretical positions such as Piagetian structuralism and dialecticism. 156 FOOTNOTES PART ONE Chapter 1: 1. Although he did not state so ex p l i c i t l y , Lippman seemed to hold a theory of the self as a system of hierarchical stages of awareness, the core self being that unit in the system which can monitor the cognitive activities of the lower order 'stages'. This perspective, of course, is entirely consistent with a neo-Platonic theoretical orientation. Chapter 2: 2. A more detailed version of this chapter was presented to the annual meeting of the Canadian Psychological Association in June, 1976, and was entitled: "A Methodological note on the measurement of personal and social stereotypes: Check l i s t s versus an open-ended format". PART TWO Chapter 4Y 3. The respondents were asked also to go back over their set of "ticked" words and to mark with a star(*) the ten tr a i t s wich seemed to them to be the most typical of the targets in question. However, since these particular responses were not examined in the data analyses they are being excluded from the main discussion of the materials and procedure. 4. The words were selected from the social stereotypes of the Americans (a positively perceived group), the Germans ( a target in the present study) and the Turks (an unfavourably perceived group) on the basis that 157 they were not similar in meaning to the 60 stimulus words. They thus offered a choice of terms with which to characterize the Orkney Islanders that extended beyond the items in Table 1. 5." Webster Is New Dictionary of Synonyms (1968) and The New Webster  Encyclopedic Dictionary of the English Language (1971) were used when scrutinizing the l i s t for redundant and semantically overlapping terms. The items in the present l i s t were drawn from one pool of terms created in 1964 (Anderson, 1968) and another created as long ago as 1933 (Karlins, Coffman and Walters, 1969). Thus the temporal validity of our instrument warrants consideration. Fortunately, the data at hand are reassuring. The 'likeableness' and meaningfillness' ratings of the words in the Anderson l i s t have proven to possess a high degree of r e l i a b i l i t y across diverse groups of male and female student respondents (Anderson, 1968). Furthermore, Hartsough and Fontana (1970) reported that a group of collaborators indicated that 86% of the original Katz and Braly (1933) l i s t of adjectives were s t i l l relevant in describing social stimuli more than a quarter of a century after they were f i r s t used. Consequently, the question concerning the long term 'meaningfulness' of the list e d items does not appear to be an important one. 6. An additional piece of information was sought which I thought might prove interesting, but by and large the instructions were misinterpreted and no useful findings emerged. The request was as follows: Write two or three sentences on how you think your ideas about different social,rregional, and national groups are formed. 158 7. The person's age in years rather than in years and months was taken as the unit of analysis for ANOVA. 8. The "explanation" was included because many respondents asked about the relationship between the two testing sessions as they entered the " f i e l d laboratory" for the second time. 9. For example, eighty-nine (46%) of the one hundred and ninety-three respondents had one or more words in common between their free response and check l i s t descriptions of the Orkney Islanders. Of these, sixty-nine (36%) had one, fifteen (8%) had two, and four (2%) had three overlapping terms. In a l l , only eighteen of the seventy five check l i s t t raits appeared among the free response items. The absence of a great deal of commonality between the items provided by the free response and structured measuring techniques reinforces the argument (Chapter 2) that the instruments tap complementary components (the core and the periphery) of personal stereotypes. 10. A number of the respondents failed to provide favourableness ratings for some of the terms in their personal stereotypes of the Orkney Islanders. However, i t was possible to estimate quite a few of these missing values by averaging the ratings attributed to the items in question when they were used in the respondents' characterizations of the other three targets. When i t was not possible to estimate missing values the favourableness scores were computed from the incomplete sets of available ratings. There was a small number of more extreme 159 cases, where descriptions lacked ratings whatsoever. In these cases, a neutral score of 5.0 was ascribed to the characterizations The same problems occurred and the same resolution was adopted with respect to the participants' estimations of the Orkney Islanders' concepts. 11. In some instances ties occurred for tenth position on the l i s t of most frequently used words. Random selection was employed in these cases to determine which terms would remain and which would be deleted from the shortlist of ten social stereotype items. 12. When obtaining these scores i t became necessary to categorize the {relatively few) items which had not received a favourableness rating. The following procedure was adopted. The respondents' own personal stereotypes of the targets and their estimations of how the Orkney Islanders view the target groups were treated separately. When a term had not been used to characterize any of the other targets and consequently could not be classified on the basis of i t s ratings in those other contexts i t was counted as being neutral. When i t had been rated one other time i t was categorized on the basis of that rating. When i t had been used and rated in the descriptions of two other targets i t was classified with respect to the more favourable value.(In general a majority of the words in the individuals'' descriptions were favourable and thus a more favourable rating could be considered more normative). Finally, a tra i t that had been used to describe a l l of the other three targets and that had been rated three times in the process was categorized according to the modal rating (e.g., four in the set of values 4,4, and 160 2) or the most frequently used segment of the scale (e.g. positive when 7,8, and 5 were l i s t e d ) . Chapter 5: 13. This finding was supported by the responses to two items in the post-film questionnaire. When asked: Do you know more about the Orkney Islanders after watching the film ? — Eighty percent (151) of those who answered did so in the affirmative. (13% said "no" and 7% "couldn't say"). When asked: How much more do you know ? — Forty one percent (64) said "some", and 21% (32) responded - "a l o t " . 14. The "two week test interval" subgroup was hot compared with the officer candidate subgroup because the members differed along a number of demographic and role dimensions. Any group differences that might have been detected could be attributed to these factors just as readily as they could be explained by time-lag effects. 15. Since the ns for the treatment conditions differed somewhat, the degrees of freedom for the tests were based on the largest n_. This procedure results in a positively biased test, in that the hypothesis of homogeneity is rejected more frequently than would be the case normally. However, there were no instances in which significant outcomes would have been reversed with fewer degrees of freedom. 16. The homogeneity of covariance assumption was not tested since the heterogeneity of variances and the unequal sample sizes were considered to be sufficient grounds for adopting a conservative approach in the data analyses. 17. The differentiation scores for the estimated reciprocal personal 161 stereotypes of Canadians were analysed together with, the scores for the personal stereotypes of the target group in a 3x2x2 between-within analysis of variance. The three-level, between-groups factor represented the different experimental conditions. The two-level, repeated measures factors represented the stereotypes of the Orkney Islanders versus the role-played stereotypes of Canadians, and the core (free response) versus the peripheral (check l i s t ) components of these stereotypes, respectively. The between-groups main effect was non-significant (F_ £2,190)= 2.78, £<i07). The estimated stereotypes were no less differentiated than the personally endorsed characterizations (F (1,190)= 2.44, p_ = .12) and there was no s t a t i s t i c a l l y significant interaction between groups and type of stereotype (F (2,190)=22.84, £ = .06). 18. Two points are worth noting when considering the commonness scores. A 3x2x2 between-within, unweighted means analysis of variance (see footnote #16 for details about the independent variables) yielded a significant difference between groups (J_ (2,190)= 11.60, £<.001). It also Indicated that the estimated reciprocal stereotypes were more idiosyncratic and contained less commonly used terms than the personally endorsed stereotypes of the Orkney Islanders (F(-l,190) = 33.18, £ < .0005). 19. The positive reciprocal group members listed a total of 199 different items in their open-ended descriptions of the Orkney Islanders, the non-reciprocal respondents listed 177, and the negative reciprocal group members listed 181 t r a i t s . In addition, a l l of the 75 check l i s t terms were used by one or more of the individuals in the positive and negative 162 reciprocal conditions, while 73 out of the 75 were used by the non-reciprocal group members when characterizing the Orkney Islanders (unappreciative and intolerant were the excluded Items). 20. The pattern of mean differentiation scores In Table 11 suggests that the age differences were not confounded with stimulus material variations in determining the degree of differentiation of the core components of the respondents' personal stereotypes. If there had been such a confound i t would have been reflected in an experimental group x stereotype component interaction in which the mean differentiation score for the core components of the negative reciprocal group members' characterizations was low relative to the mean score for the peripheral components. 21. It is worth mentioning at this point that the multiple comparisons were also conducted with Scheffe's (1953) procedure and the test outcomes corroborated those provided by the t^ s t a t i s t i c . 22. Because of the large number of tests performed, results that were significant at the .05 level were treated with caution and were discussed only when a meaningful interpretation was clearly apparent. 23. The Fs for the differences between stereotype components regarding the number of favourable and neutral words that were included are 206.26, p_<.0005 and 24.47, p_<.0005 for the positive reciprocal group; and 91.80, £,£.0005 and 27.13, £<.0005 for the negative reciprocal group. The degrees of freedom and 1 and 190 in each case. The Fs for the differences between the core and peripheral components regarding their inclusion of unfavourable terms are 1.09 for the positive and 0.29 for the non-reciprocal treatment conditions, respectively. In addition^ 163 the Fs for the differences between- the non^-reciprocal group members' stereotype components regarding the number of favourable and neutral words that were included are 78.73, p_<.0005 and 58.69, p_<.0005, respectively. 24. The Fs, with 1 and 190 degrees of freedom, for the differences between the number of favourable, neutral, and unfavourable words in the peripheral components of the personal stereotypes are 118.03, p_<.0005 for the positive reciprocal condition; 48.37, p_<.0005 for the non-reciprocal group; and 29.35, p_<.0005 for the negative reciprocal treatment condition. The equivalent Fs for the core components are 6.27, p_<.05; 2.70, ns_; and 1.02, ns, respectively. The t^ tests were conducted using conservative c r i t i c a l values. 164 BIBLIOGRAPHY Abate, M., & Berrien, F. K. Validation of stereotypes: Japanese versus Americans:students. Journal of Personality and Social Psychology, 1967, ]_, 435-438. Aboud, F.E. The effect of stereotype generalization on information seeking. Canadian Journal Of Behavioural Science, 1976, 8_, 178-188. Aboud, F.E., & Taylor, D.M. Ethnic and role stereotypes: Their relative importance in person perception. Journal Of Social Psychology, 1971, £[r85_7-17r27. Aboud, F.E., t. Taylor, D.M., & Doumani, R.G. The effect of contact on the use of role and ethnic stereotypes in person perception. Journal of Social Psychology, 1973, 89, 309-310. Anderson, N.H. Primacy effects in personality impression formation using a generalized order effect paradigm. Journal of Personality and Social Psychology, 1965, 2_, 1-9. Anderson, N.H. Likableness ratings of 555 personality-trait words. Journal of Personality and Social Psychology, 1968, 9_, 272-279. Asch, S.E. Forming impressions of personality. Journal of Abnormal and \ Social Psychology, 1946, 41_, 258-290. Avigdor, R. Etude experimentale de l a genese des stereotypes. Cahiers  Internationaux de Sociologie, 1953, 14, 154-168. 165 Bailey, P. Orkney. Newton Abbot: David and Charles, 1971. Bank Ahead. Film: National Commercial Bank of Scotland, 1967. Bayton, J.A. The racial stereotypes of Negro college students. Journal of Abnormal arid Social Psychology, 1941, 36_, 97-102. Bayton, J.A., & Byoune, E. Racio-national stereotypes held by Negroes. Journal of Negro Education, 1947, 1_6, 49-56. Bayton, J4:Ai\, McAlister, L.B., & Hamer, J. Race-class stereotypes. Journal of Negro Education, 1956, 25_, 75-78. Berenson, B. The central Italian painters of the Renaissance. 1897. ,JReprinted in Berenson, B. The Italian painters of the renaissance. Oxford: Oxford University Press, 1932. Birnbaum, M.H. Morality judgements: Tests of an averaging model. Journal of Experimental Psychology, 1972, 93, 35-42. Bjerstedt, A. "Ego-involved world-mindedness," nationality images, and methods of research: A methodological note. Journal of Conflict " Resolution, 1960, 4_, 185-192. Blake, R., & Dennis, W. The development of stereotypes concerning the Negro. Journal ofdAbnormal and Social Psychology, 1943, 38, 525-531. Bogardus, E.S. Stereotypes versus sociotypes. Sociology and Social ?Research, 1950, 34_, 286-291. Brigham, J.C. Ethnic stereotypes. Psychological Bulletin, 1971, 7_6, 15-38. Brigham, J.C. Ethnic stereotypes and attitudes: A different mode of analysis. Journal of Personality, 1973, 41, 206-223. 166 Brovermati, L.K., Broverman, D.M., Clarkson, F.E., Rosenkrantz, P.S., & Vogel, S.R. Sex-role stereotypes and c l i n i c a l judgements of mental health. Journal of C l i n i c a l arid 'Consulting Psychology, 1970, 34, 1-7. Brown, R. Words arid things. Glencoe, I l l i n o i s : The Free Press, 1958. Brown, R. Social psychology. New York: The Free Press, 1965. Bruner, J.S., Shapiro, D., & Taglurl, R. The meaning of traits i n isolation and combination. In R. Tagiuri and L. Petrullo (Eds.), Persdri perception arid interpersonal behaviour. Stanford, California: Stanford University Press, 1958. Cahalan, D., & Trager, F.N. Free answer stereotypes and anti-Semitism. Public Opinion Quarterly, 1949, 13, 93-104. Campbell, D.T. Stereotypes and the perception of group differences. American Psychologist, 1967, 22_, 817-829. Clark, K.B., & Clark, M. P. Skin colour as a factor in racial identification of Negro pre-school children. Journal of Social Psychology, 1940, 11_, 159-169. Collier's Encyclopaedia. Crowell-Collier Education Corporation, 1972. Cook, M. Interpersonal perception. Harmondsworth: Penguin Education, 1971. Cooley, W.W., & Lohnes, P.R. Multivariate data analysis. New York: John Wiley & Sons, 1971. Curtis, L.P., Jnr. Apes and angels: The Irishman in Victorian Caricature. Newton Abbot: David and Charles, 1971. Cusumano, D.R., & Richey, M.H. Negative salience in impressions of character: Effects of stimulus information. Psychonomic Science, 1970, 20, 81-83. 167 Diab, L.N. National stereotypes and the \eference group" concept. Journal of Social Psychology, 1962, 57_, 339-351. Dudycha, G.L. The attitudes of college students toward war and the Germans before and during the Second World War. Journal of Social Psychology, 1942, 15, 317-324. Duijker, M.C.J., & Frijda, N.J. National Character and national ,„ stereotypes: A trend report prepared for the International Union of Scientific Psychology. Amsterdam: North-Holland, 1960. Ehrlich, H.J., & Rinehart, J.W. A brief report on the methodology of stereotype research. Social Forces, 1965, 43, 564-575. E l l i s , L.J., & Bentler, P.M. Traditional sex-determined role standards and sex stereotypes. Journal of Personality and Social Psychology, 1973, 25, 28-34. Encyclopaedia Britannica. Chicago: Encyclopaedia Britannica Inc. and Benton, 1968. Eysenck, H.J., & Crown, S. National stereotypes: An experimental and methodological study. International Journal of Opinion and Attitude  Research, 1948, 2, 26-39. Feldman, J.M.,Stimulus characteristics and subject prejudice as determin-ants of stereotype attribution. Journal of Personality and Social  Psychology, 1972, 21, 333-340. Feldman, J.M., & Hilterman, R.J. Stereotype attribution revisited: The role of stimulus characteristics, racial attitude, and cognitive differentiation. Journal of Personality and Social Psychology, 1975, 31_, 1177-1188. 168 Fishman, J.A. An examination, of the process and function, of social stereotyping. Journal of Social "Psychology, 1956, 43, 27-64. FunktaridsWagrialls Standard Handbook of Synonyms,"Antonyms, and Prepos- itions. New York: Funk and Wagnalls, 1947. Games, P.A. Inverse relation between the risks of Type 1 and Type 2 errors and suggestions for the unequal n_ case in multiple comparisons. Psychological Bulletin, 1971, 75, 97-102. Gardner, R.C. Ethnic stereotypes: The traditional approach, a new look. Canadian Psychologist, 1973, 14_, 133-148. Gardner, R.C, Kirby, D.M., Gorospe, F.H., & Villamin, A.C. Ethnic stereotypes: An alternative assessment technique, a stereotype - di f f e r e n t i a l . Journal of Social Psychology. 1972, 8_7, 259-267. Gardner, R.C, Kirby, D.M., Pablo, R.Y., & Castillo, E.S. Ethnic stereotypes; The role of language. Journal of Social Psychology, 1975, 96, 3-9. Gardner, R.C, Rodensky, R.L., & Kirby, D.M. Ethnic stereotypes: A  c r i t i c a l review. University of Western Ontario, Research Bulletin no. 157, 1970. Gardner, R.C, Taylor, D.M., & Feenstra, H.J. Ethnic stereotypes: •/Attitudes or beliefs ? Canadian Journal of Psychology, 1970, 24, .321^334. Glass, G.V., & Stanley, J.C Sta t i s t i c a l methods in education and  psychology. Englewood C l i f f s , New Jersey: Prentice-Hall, 1970. Gilbert, CM. Stereotype persistence and change among college students. Journal of Abnormal and Social Psychology, 1951, 46, 245-254. 169 Goodman, M.E. Race awareness in young children. Cambridge, Mass.: Addison- Wesley, 1952. Gray-Little, B. The salience of negative information i n impression formation among two Danish samples. Journal of Cross-cultural Psychology, 1973, 4_, 193-208. Haque, A. Stereotypes of college students in Hyerabad toward the 1. ingroup and the outgroup during the Indo-Pakistan war. Journal of Social Psychology, 1968, 74_, 285-286. Haque, A. National stereotypes of college students in Hyerabad during the . Indo-Pakistan war. Pakistan Journal of Psychology, 1969, 2_, 21-38. Hamilton, D.L., & Huffman, L.J. Generality of impression formation processes for evaluative and nonevaluative judgements. Journal of Personality and Social Psychology, 1971, 20-, 200-207. Hamilton, D.L., & Zanna, M.P. Differential weighting of favourable a.rand unfavourable attributes in impressions of personality. Journal of Experimental Research in Personality, 1972, <6, 204-212. Hartsough, W.R., & Fontana, A.F. Persistence of ethnic stereotypes and the relative importance of positive and negative stereotyping for association ./preference. Psychological Reports, 1970, 27_, 723-731. Hayakawa, S.I. Recognizing stereotypes as substitutes for thought. Etc.; A Review of General Semantics, 1950,'.'^  208-210. Hicks, J.VM., Goldman, M., & Kang, J. Attitudes toward Negroes and stereotypes about Americans among Chinese students in Taiwan and the United States. Journal of Social Psychology, 1968, 76_, 139-141. 170 Hudson, L. Contrary Imaginations: a psychological study of the English schoolboy. Harmondsworth: Penguin Books, 1967. Hudson, L. Frames of mind: Ability, perception and self-perception in the arts and sciences. Harmondsworth: Penguin Books, 1970. Insko, CA., & Robinson, J.E. Belief similarity versus race as determinants of reactions to Negroes by Southern white adolescents: A further test of Rokeach's theory. Journal of Personality and Social Psychology, 1967, 1,216-221. Jones, E.E., & Davis, K.E. From acts to dispositions: The attribution process in person perception. In L. Berkowitz (Ed.) Advances in experimental social psychology. Vol.2. New York: Academic Press, 1965. Jones, R.A., & Ashmore, R.D. The structure of intergrouppperception: Categories and dimensions in views of ethnic groups and adjectives used in stereotype research. Journal of Personality and Social Psychology, 1973, 25, 428-438. Johnson, R.E., & Downing, L.L. Attribution of a b i l i t y as a function of consistency of information across times, tasks, and persons: A bias tow-I...;ard favourable attributions. Psychological Reports, 1976, 38_, 99-105. Kanouse, D.E., & Hanson, L.R., Jnr. Negativity in evaluations. (General Learning Press Module Number 4013V00) Morristown, N.J.: General Learning Press, 1972. Kaplan, R.M., & Goldman, R.D. Interracial perception among Black, white, and Mexican-American high school students. Journal Of Personality and Social Psychology, 1973, 28, 383-389. Karlins, M., Coffman, T.L., & Walters, G. On the fading of social stereotypes: Studies in three generations of college students. Journal of Personality and Social Psychology, 1969, 13, 1-16. 171 Katz, D., & Braly, KW. Racial stereotypes in one hundred college students. "Journal of Abnormal arid Social Psychology, 1933, 28, 280-290. Katz, D., & Braly, K.W. Racial prejudice and racial stereotypes. Journal of Abnormal and Social Psychology, 1935, 30_, 175-193. Kelvin, P. The bases of social psychology. New York: Holt, Rinehart, and Winston, 1970. Kendon, A., & Cook, M. The consistency of gaze patterns in social inter-action. British Journal of Psychology, 1969, 60, 481-494. Kirby, D.M., & Gardner, R.C. Ethnic stereotypes: Determinants i n children and their parents. Canadian Journal of Psychology, 1973, 27, 127-143. Kirk, R.E. Experimental desigri: Procedures for the behavioural sciences. Belmont, California: Brooks/Cole, 1968. Klineberg, 0. The human dimension in international relations. New York: Holt, Rinehart and Winston, 1964. Koltuv, B. Some characteristics of intrajudge t r a i t intercorrelations. Psychological Monographs, 1962, 76, Whole number 552. La Gaipa, J.J. Stereotypes and perceived ethnic-role specialization. Journal of Social Psychology, 1971, 8_5, 285-292. Lambert, W.E., & Klineberg, 0. A pilot study of the origin and development of national stereotypes. International Social Scierice Jourrial, 1959, U, 221-238. LaPIere, R.T. Type-rationalizations of group antipathy. Social Forces, 1936, 15, 232-237. 172 Lay, C.H., & Jackson, D.N.. Analyses of the generality of t r a i t -inferential relationships. "Journal of Personality and Social ^Psychology, 1969,12, 12-21. Lippmann, W. Public Opinion. New York: Harcourt, Brace, 1922. 17th. printing in 1961. Litterer, O.F. Stereotypes. Journal of Social Psychology, 1933, 4_, 59-69. Marston, B.J. Trait attribution to a described action as a function of changes i n salient information. Journal of Re. search in Personality, 1976, 1_0, 245-255. Maykovich, M.K. Changes in racial stereotypes among college students. Human Relations, 1971, 24, 371-386. Maykovich, M.K. Changes in racial stereotypes among college students. British Journal of Social Psychiatry and Community Health, 1972, 6_, 126-133. McKeachie, W.J. Lipstick as a determiner of f i r s t impressions of personality. Journal of Social Psychology, 1952, 36, 241-244. McTiernan, T.J.;; The Gael and the Gall: A study of national and subnational stereotypes. Unpublished B.A. thesis, Trinity College, v Dublin, 1973. Meenes, M. A comparison of racial stereotypes of 1935 and 1942. Journal of Social Psychology, 1943, 17_, 327-336. Morrissette, J. An experimental study of the theory and structure of balance. Human Relations, 1958, 11, 239-254. 173 Norman, W.T. Toward an adequate, taxonomy of personality attributes; Replicated factor structure in peer nomination personality ratings. Journal of Abnormal arid Social Psychology, 1963, 66_, 574-583. O'Hara, R. Some implications of stereotyping, In I. Deer and H. Deer (Eds.) C - c ) Lariguages of the mass media. Boston: D.C. Heath, 1965. Osgood, C , Sucl, G.J., & Tannenbaum, P.H. The measurement of meaning. Urbana, I l l i n o i s : University of I l l i n o i s Press, 1957. Passlnl, F.T., & Norman, W.T. A universal conception of personality structure? Jourrial of Personality and Social Psychology, 1966, 4_, 44-49. Petrinovich, R.F., & Hardyck, CS. Error rates for multiple comparison methods: Some evidence concerning the frequency of erroneous conclusions. Psychological Bulletin, 1969, 71, 43-54. Pettigrew, T.F. Personality and sociocultural factors in inter-group attitudes: a cross national comparison. In C.W. Backman and P.F. Secord (Eds.) Problems in social psychology: Selected readings. New York: McGraw-Hill, 1966. Prothro, E.T., & Melikian, L.H. Studies in stereotypes: III. Arab students in the Near East. Journal of Social Psychology, 1954, 40, 237-243. Prothro, E.T., & Melikian, L.H. Studies in stereotypes: V. Familiarity and the kernel of truth hypothesis. Journal of Social Psychology, 1955, 41, 3-10. Radke, M., Trager, H.G., & Davis, H. Social perceptions and attitude^ of children. Genetic Psychology Monographs, 1949, 40_, 327-447. 174 Reigrotski, E., & Anderson, N. National stereotypes and foreign contacts: Public Opinion Quarterly, 1959, 23, 515-528. Rice, S.A. "Stereotypes": A source of error in judging human character. Journal of Personnel Research, 1926, _5, 267-271. Richey, M.H., & Dwyer, J.D. Negative salience in impressions of character: Sex differences. Psychbnomic Science, 1970, 20, 77-79. Richey, M.H., Koenig, R.J., Richey, H.W., & Fortin, R. Negative salience in impressions of character: Effects of unequal proportions of positiveand negative information. Journal of Social  Psychology? 1975, 97, 233-241. Richey, M.H., McClelland, L., & Shimkunas, A.M. Relative influence of positive and negative information in impression formation and persistence. Journal of Personality and Social Psychology, 1967, 6^  322-327. Riskey, D.R., & Birnbaum, M.H. Compensatory effects in moral judgement: Two rights don't make a wrong. Journal of Experimental Psychology, 1974, 103, 171-173. Rokeach, M. The Open and Closed Mind« New York: Basic Books, 1960. Rokeach, M., Smith, P.W., & Evans, R.I. TTwo kinds of prejudice or one ? In Rokeach, M. The Open and Closed Mind. New York: Basic Books, 1960. Saenger, G., & Flowerman, S. Stereotypes and prejudicial attitudes. Human Relations, 1954, 7_, 217-238. Scheffe, H. A method for judging a l l contrasts in the analysis of variance. Biometrika, 1953, 40, 87-104. Schuman, H. Social change and the validity of regional stereotypes in East Pakistan. Sbciometry, 1966, 29_, 428-440 175 Seago, D.W. Stereotypes: before. Pearl. Harbor and after. Journal of Psychology, 1947, 23, 55-63. Secord, P.F., & Backman, C.W. Social Psychology. New York: McGraw-H i l l , 1964. Sherif, M., Harvey, O.J., White, B.J.,Hood, W.R. & .Sherif, -C^ :^ V Intergroup conflict and cooperation: the Robbers Cave experiment. Norman,Oklahoma: The University of Oklahoma Press, 1961. Si g a l l , M., & Page, R. Current stereotypes: a l i t t l e fading, a l i t t l e faking. Journal of Personality and Social Psychology, 1971, 18, 247-255. Signori, E.I., & Butt, D.S. Ratings of the social images of disadvantaged groups by males and females. Psychological Reports, 1972, 3_0, 575-580. Simmons, O.G. Stereotypes: Explaining people who are different. In Spradley, J.P., & McCurdy, D.W. (Eds.) Conformity and Conflict: Readings in cultural anthropology. Boston: L i t t l e , Brown, 1971. Simpson, G.E., & Yinger, J.M. Racial and Cultural minorities. (Rev.ed.) New York: Harper and Row, 1965. Sinha, A.K.P., & Upadhyay, O.P. Stereotypes of male and female university students in India toward different ethnic groups. Journal  of Social Psychology, 1960, 51, 93-102. (a) Sinha, A.K.P., & Upadhyay, O.P. Change and persistence i n the stereotypes of university students toward different ethnic groups during the Sino-Indian border dispute. Journal of Social Psychology, 1960, 52, 31-39. (b) Smith, M.B., Bruner, J.S., & White, R.W. Opinions and personality. New York: Wiley and Sons, 1956. Stanley, G. Australian students' attitudes to Negroes and Aborigines 176 on the Multifactor Racial Attitude Inventory (MRAI). Journal of Social Psychology, 1969,^ 77_, 281-282. Stein, D.D., Hardyck, J.A., & Smith, M.t:\,B. Race and belief: An open and shut case. Journal of Personality and Social Psychology, 1965, J_, 281-290. S t i l l w e l l , R.j & Spencer, C. Children's early preferences for other nations and their subsequent acquisition of knowledge about those nations. European Journal of Social Psychology, 1973, 3_, 345-349. Strongman, K.T., & Hart, C.J. Stereotyped reactions to body build. Psychological Reports, 1968, 23, 1175-1178. Sun, K. Adults and adolescents: Favorablenessaand differentiation of role stereotypes. Unpublished term paper. University of British Columbia, 1976. Szalay, L.B., & Maday, B.C. Verbal associations In the analysis of subjective culture. Current Anthropology, 1973, 14, 33-42. Tagiuri, R. Social preference and i t s perception. In Tagiuri, R., & Petrullo, L. (Eds.), Person Perception and interpersonal behaviour. Stanford, California: Stanford University Press, 1958. Tagiuri, R., Blake, RY'R., & Bruner, J.S. Some determinants of the perceptionsof positive and negative feelings in others. Journal  of Abnormal and Social Psychology, 1953, 48_, 585-592. Tanaka, Y. A p o l i t i c a l d i f f e r e n t i a l , or POLDI, for the measurement of national images. Unpublished research, Gakashuin University, Tokyo, 1965. Tanaka, Y. A study of national stereotypes. In Triandis, H.C., 177 Vassiliou, V., Vassiliou,G., Tanaka, Y., & Shanmugam, A.V. The analysis of subjective culture. New York: Wiley-Interscience, 1972. Taylor, D.M., & Aboud, F.E. Ethnic stereotypes: Is the concept necessary? Canadian Psychologist, 1973, JL4, 330-338. Taylor, D.M., B a s s l l i , iT.N., & Aboud, F.E. Dimensions of ethnic Identity: An example from Quebec. Journal of Social Psychology, 1973, 89, 185-192. The New American Rbget's Thesaurus In Dictionary Form. New York: Signet, 1962. The New Webster Encyclopedic Dictionary of the English Language. Chicago: Consolidated Book Publishers, 1971. Thorngate, W. Process invariance:-another red herring. Personality and Social Psychology Bulletin, 1975, 1_, 485-488. Triandis, H.C., & Davis, E.E. Race and belief as determinants of behavioural intentions. Journal of Personality and Social Psychology, 1965, 2_, 715-725. Vinacke, W.E. Stereotypes as social concepts. Journal of Social Psychology, 1957, 46, 229-243. Webster's New Dictionary of Synonyms. Springfield, Mass.: G.& C. Merriam, 1968. Weinstein, E.A., & Crowdus, S.E. The effects of positive and negative information on person perception. Human Relations, 1968, 21, 383-391. Wiggins, J.S. In defense of tr a i t s . Invited address to the Ninth Annual Symposium on Recent Developments in the use of MMPI, held in Los Angeles on February, 28, 1974. Winer, B.J. St a t i s t i c a l principles in experimental design, 2nd, ed. New York: McGraw-Hill, 1971. Wishner, J. Reanalysis of "Impressions of personality". Psychological Review, I960; "67, 96-112. Zajonc, R.B. The attitudinal effects of mere exposure. Journal of Personality and Social Psychology, Monograph Supplement, 1968, Part 2, 1-27. Zajonc, R.B., & Burnstein, E. Structural balance, reciprocity, and positivity as sources of cognitive bias. Journal of Personality, 1965, 33, 570-583. 179 APPENDIXES APPENDIX A: SCRIPTS FOR THE TAPE RECORDED DOCUMENTARY PRESENTATION 181 Each of the three documentary scripts f a l l into two parts. The f i r s t section was common to a l l of the conditions and the second contained one of the three versions of the experimentally manipulated information. 182 Common portion of script "Fifty miles south of the latitude of Cape Farewell at the southern tip of Greenland, and level with Churchill on Hudson Bay and Skagway in Alaska ," the Orkney Islands l i e bottled between the North Sea and the Atlantic. Separated from Norway by hundreds of miles of cold sea and from the northern tip of Scotland by racing, treacherous, unpredictable r i p -currents, and ti d a l races, the islands stand remote and alone, almost s e l f -contained . Yet, despite their relative isolation, the f e r t i l e Orkney Islands have been occupied since the second ice-age. Indeed, these meticulously crafted, stone-wall houses of Kirkwall, the principal town on the islands, reflect on architectural tradition that goes back beyond recorded history, back three thousand years to the mystery which surrounds the neolithic sunken village of Skara Brae on the shores of Scapa Flow and the circles, of standing stones which dot the islands as re l i c s of some unknown and unrecorded race. Walking towards Kirkwall harbour, these two men, one a bank accountant, the other a sailor, are typical of the quiet, conventional, meticulous Orkney Islanders. Linked together by their unique jobs, they portray that blend of the modern with the traditional that Is so common to l i f e on the Orkneys. "William Groat, banker, and Davey Irving, skipper, have been running a branch bank for years*5." Their branch office i s l i k e no other. Theirs i s a boat, and they take to the sea, carrying out twentieth century 183 monetary transactions i n a tradition that has not changed much since the Viking longboats sailed throughout this island archipelago, a thousand years ago. "The National Commercial Bank of Scotland bought a boat in 1952c." Following the sea-passages charted by the agressive, irreligious and somewhat ungraceful forefathers of the modern islanders, they set out to serve the people who were too isolated and too busy to make the long trip to the banks of Kirkwall and Stromness, the only two towns on the island. The idea of a floating bank, serving the remote communities of the outlying islands, has been seen by some to point to a certain daydreamer characteristic of the islanders. Eccentric daydreamers though they may be, their persistent vision and adaptation to changing economic conditions have meant that both farmers and small businessmen on the islands can compete and survive in the face of strong southern competition. Like most boats on the islands the Otter Bank is strong and sea-worthy, and i s captained by an experienced seaman. Indeed, i t i s essential to the islanders that they have many good sailors. The sea is their major highway and i t can be fickle and dangerous. Nevertheless, unlike the Canadian Indians of the Pacific Northwest coast who also use the sea as a highway, Orcadians are not sailors i n the traditional sense. Although the sea influences their lives they do not look to i t for a living. They have no large fishing fleet, nor do many of the men go to sea for a long period of time. Farming i s their way of l i f e . 184 Farmers though they may be, the winds and tides and the raging storms have had their effects on the islanders' characters. They appear blunt. Their bluntness, however, cloaks a proud, fearless quality which is mixed with a meditative and self-possessed sentimentality. They care for the past and worry about the future of the islands. They are determined that depopulation and transportational d i f f i c u l t i e s w i l l not strangle their close-knit agricultural communities. "The floating bank's parent ship has been anchored in a busy street in Kirkwall for over a hundred years'*," serving a town whose population is s t i l l under 5,000 people. "Kirkwall," according to one recent writer, " i s not only the capital of Orkney. It also brings together in i t s streets and buildings a l l the history and a l l the l i f e of Orkney, past and present ." Until the growth of Stromness, some two hundred years ago, i t remained the only town in the sixty or more islands that form the archipelago. The agricultural orientation of the islanders i s reflected in the businesses and appearance of Kirkwall. As Patrick Bailey, the writer, puts i t : "There are relatively few signs that Kirkwall i s a seaport, many that i t i s the centre of a farming community. Even the Customs House also accommodates the islands' College of Agriculture." The bank manager of the National Commercial i s an Orcadian by birth. He feels that he has a particular advantage over many of his fellow managers i n that he has never worked in any bank other than the one i n Kirkwall. As a result, he knows his customers intimately. He knows their families and friends. As an islander, he can help them with their problems 185 and sympathize with them in their adversity. Most importantly, the people can know that, as a neighbour, he Is genuine and concerned i n his feelings. One can find mixed opinions about banking among the towns-people. While one lady might feel that she has benefited from opening a deposit account, there are others who would disagree. One man, who was inter-viewed, f e l t that his money was safer under the bed, another woman con-sidered her income to be too small to be of interest to the bank. The islanders are c r i t i c a l and naive by nature. It takes time to convince them that certain changes can be useful and beneficial. However, influence is exerted by such people as a hotelier who advises her staff to bank their tips and build up a savings account, or by Boyce Swaney, the bank manager, whose easy mannerisms and approach-ab i l i t y allow people to be relaxed and comfortable in his presence. He understands that many traditional beliefs and customs have to be overcome before he can hope to serve the entire community and not just the businessmen and farmers. As bank manager he expresses optimism for the future of banking and commerce in Orkney in the same breath that he argues for the preservation of the "sacred" customs of the past. In this s p i r i t of careful and planned change, more and more of the islanders are overcoming their doubts and qualms and are adding their names to the l i s t of bank customers as they grown to realize that the future of the islands l i e s i n their hands and can be based only on a thriving economy centred around an active participation in investment and business. 186 And as the islands' population of 18,000 people modify their values and attitudes to f i t twentieth century demands, like bank manager Boyce Swaney, they s t i l l hold dear to many of the old traditions. The buying and selling of cattle has taken a new form since the days of Lammas Fair, which took place once a year and lasted for two weeks. Even the old market place on Kirk Green, in front of the twelfth century Cathedral of St. Magnus has been abandoned for new and modern quarters. Yet the men, just as they have always done, s t i l l stand close-packed and silent around the cattle pens making their bids quietly and inoffen-sively i n that moderate manner that i s so common to a l l Orcadians. As a farming community, beef-cattle provide a major source of revenue for the Orkney islanders. "Livestock numbers are approximately 70,000 beef cattle, 6,000 dairy cattle, 78,000 sheep and a small number of pigs^." "Despite the remoteness and heavy shipping charges Orkney farmers have prospered through hard work and shrewd investment^." Almost a l l of the holdings are worked as family farms, the majority ranging i n size between 50 and 150 acres. Many of these farms have been built from smaller units that were originally created in previous centuries under a more restrictive agricultural policy. George Rauss is an Englishman who bought a farm, through the bank, while he was s t i l l i n the army. An outsider, he i s not fully a part of the inner c i r c l e of the Orcadian community. Nevertheless he has integrated well, and shares many strong friendships with his neighbours. 187 Like his neighbours and fellow farmers he depends on the bank to help him in his agricultural a c t i v i t i e s . In a rural community which lives on seasonal earnings the bank has replaced the small businesses and wholesalers who can no longer extend credit u n t i l the spring and f a l l livestock sales. In an age of spiralling costs, feed and f e r t i l i z e r must be purchased as cheaply as possible from large companies which are subsidiaries of even larger corporations which w i l l not and, perhaps, cannot extend prolonged credit. The bank has stepped i n to provide annual overdrafts which w i l l allow George and his neighbours to continue their work without the worry of pressing creditors. Farming the Orkney s o i l i s hard work, but the islanders are fortunate enough to have rich deep deposits of f e r t i l e glacial clay which yield • good crops and grassland. The land is s t i l l bountiful to the present generation of Orcadians. Their beef i s highly prized and their sheep, fed largely on sea-weed and the coarse grasses of the foreshore, have a delicate and subtle flavour that i s much in demand. Like l i f e throughout the Orkney islands farming is a blend of the traditional and the new with changes coming about slowly and carefully and always with a view to preserving the best traditions of the past. While the harvests are s t i l l gathered by groups of friends and neighbours, in keeping with the old customs, modern machinery i s beginning to appear in the fields. Egg producing, which has developed rapidly since the second world war, has always been a forerunner among Ordadian enterprises in terms of modernization and expansion. 188 It required vast amounts of capital in addition to confident foresight in order to enter a unknown and competitive market. The Orkney islanders weighed the odds and took the plunge. It says much for their fearlessness and persistence that they have survived a severe marketing slump to pull through and start on the long road to recovery once again. At present, "nearly one million eggs leave Orkney every week for the markets of the south h." In exporting their produce, be i t eggs, cattle, cheese, or some more exotic delicacies, the Orkney Islanders are faced once again with the sea, and with the need to use i t as a route of transportation. "Steamers c a l l at most of the larger islands nearly twice a week. Yet, this does l i t t l e to dispel the feeling of remoteness that s t i l l exists, especially in the northern-most i s l e s * . " The floating shops that used to c a l l from harbour to harbour are gone. Otter Bank alone, continues their tradition. The sea i s not central to the islanders' farming l i f e s t y l e but i t i s important and i t has had a major influence on the people of Orkney. "Otter Bank's sailings around the islands from April to September are as familiar and as regular as the mailboat's. Every Tuesday, Otter Bank heads for Shapinsay, where the children s t i l l believe that Santa Clause, l i k e the Bank, comes by boat-'." The people eagerly await i t s arrival before walking down to the small stone harbour. The local store-keeper i s considered by a l l to be a suitable lookout since his shop overlooks the entrance to the bay. His phone is constantly 189 busy as Otter Bank's estimated time of arrival draws nearer. He interrupts his work to scan the sea without feeling Imposed upon. He i s part of a small community where everyone helps everyone else. He i s the usher and master of ceremonies alike in the weekly episode that surrounds the boat's v i s i t . Many of the people who walk to to the end of the stone pier to greet the Otter Bank have no intention of doing business with banker William Groat. This i s not because they don't want to, or don't have a bank account, i t i s merely because the boat's arrival i s as much of a social event as i t i s a business t r i p . In many ways the boat i s just an excuse for people to get together and be with one another. In the case of the Otter Bank i t was not always this way. Davey Irving t e l l s about the early days of the new and innovative banking service. In ..his words: "The island people, at f i r s t , were not particu-l a r l y interested, as far as I could see, in banking. In one island especi-ally—when we arrived there the f i r s t time I remember there was not a single soul down to greet us at a l l . It's a different matter today because any time we're In there, there's always a crowd around ." Davey Irving Is not slow to t e l l people that he likes his work, especially the unofficial public relations work he does at each port of c a l l . As he says himself: "When the bank service i s being carried on i n the boat somebody has to take care of the customers who are waiting. So we generally have a yarn and pass the time of day, and get the news and share the news from Kirkwall*." 190 Some of the islanders are able to express the broad significance that Otter Bank has to the outer islands. As one man put i t : "I think there i s something more than just pounds and shillings attached to the bank's v i s i t s here, i t ' s something psychological, i t makes us feel part of an entity, part of a larger group of people™1." It i s easy to see that Davey and William are on close personal terms with the people. Their travels have been built into the fabric of l i f e on each of the islands that they v i s i t . Being from the main island they are strangers i n a certain sense, but they are familiar strangers who receive warm hospitality. The Otter Bank helps to bridge the gaps between the islands and helps to extend the sense of community from one island to the next. It brings the people into closer personal contact with the main island, with things in the outside world, and with the weekly activities in Kirkwall. There i s also a practical side to the bank's v i s i t . It takes a long time to traverse even short distances by sea. Most return journeys around the islands can take the best part of a day, some longer. Very often the farmers cannot afford to lose a day's work in order to travel to Kirkwall to carry out necessary bank transactions. In many cases, even i f they could, they would feel uncomfortable in a collar and t i e and a dark suit. It i s much better, then, from their point of view, to take off half an hour and drive down to the informal gathering at the end of the harbour where they can do business quickly and i n a relaxed manner. 191 The unorthodox, almost casual, manner in which the Otter Bank does business is reflected i n a story told with relish by both William Groat and Davey Irving. One morning the Baptist minister came to the Otter Bank in a great hurry. He had missed the mailboat to Westray and he needed to get there to perform a wedding service that evening. There is a rule that no passengers are allowed on the boat, but under such compelling circumstances the skipper and accountant f e l t compelled to look out to sea so that they need not notice the fact that the minister had climbed aboard. They made a detour to the island in question and were enticed to stay for the celebrations. Somewhat shamefacedly they admit that they had a much better time than they would have had i f they had continued on to work for the rest of the day. The Otter Bank, providing a unique service, symbolizes the neighbour-liness of the t r a f f i c between the Islands. History has seen to i t , however, that the Islanders' vision extends beyond the waters of Scapa Flow and Stronsay Firth, stretching to the world at large. Three quarters of a million barrels of whiskey a year are exported from Kirkwall's Highland Harp D i s t i l l e r y . Founded o f f i c i a l l y in 1798, i t was already in f u l l operation when the ships of the Hudson's Bay Company began to use Stromness as a regular port of c a l l . Each year, at least a hundred men and boys l e f t for service i n Canada. Indeed, the recruiting power of the Company was such that there 192 was always an imbalance between the male and female population i n Orkney's second town and largest seaport. As well as their consignment of men i t i s conceivable that many barrels of the same Highland Harp whiskey that now leaves for Sweden ended up in Canadian frontier trading posts. 1. Non-reciprocal condition Their experience with the Hudson's Bay Company has l e f t Orcadians "export minded to an astonishing degree. The world is the Orcadians' familiar market place 1 1" and Canada s t i l l proves to be the major importer of Orkney products. whether i t be whiskey, smoked salmon, or highland tweed, the great bulk of exports head toward Halifax or the St. Lawrence Seaway. As the island economy develops, however, the traditional ties with Canada are being supplemented by new and often unexpected trade routes. The island tweed, although i t is no longer hand woven, i s s t i l l of the highest quality and is much in demand in the east coast c i t i e s of the United States. There i s an arrangement between the weavers and their American agent that he sends them the colours for the coming season's fashions and they design the materials. Such an enterprise has led to the emergence of many steady jobs for skilled island workers. Present commercial ties with Canada centre around importing frozen Atlantic salmon, smoking i t , and returning i t together with large consignments of tweed and moderate amounts of whiskey. While the major Orkney industries look outward for new markets, smaller enterprises concentrate on serving the needs of the islanders themselves. A number of boat yards build vessels for the growing Orkney fishing fleet. While fishing has never been as important as i t could be in Orkney, the availability of government grants and boat building subsidies ensure that more and more of the islanders can turn away from farming and look toward the sea as a source of properous income. The future of Orkney looks bright. But i t has not always been that way. For hundreds of years, unfair land laws, absentee land-owners, and excessive rents made the islanders poor and unable to provide themselves with a high standard of l i v i n g . While magnificent buildings l i k e the Cathedral of St. Magnus were being built around them, they, themselves, lived i n poverty. Fortunately, things have changed. Land law reforms and a more understanding attitude by the outside world have allowed the islanders to forge ahead to the level of development which they have achieved today. They do not intend to s t a l l or slide backward. In the past they have been given some help i n their development by the presence of the British Navy in Scapa Flow. Now that the navy has departed that rush of quick wealth has been replaced by more stable and expanding economic markets. New dairy industries are growing up. Elec t r i c i t y and running water are coming to even the most remote of the Islands. 194 The library i n Kirkwall has the most extensive l i s t of readers in the public library system of the United Kingdom. Eighty percent of the population are regular borrowers. Orkney i s an alive and thriving community. The standard of liv i n g of the islanders i s improving rapidly. And while, as islanders, they can never f u l l y overcome the problems that arise through being separated from the mainland, their previous d i f f i c u l t i e s with transportation are growing fewer and fewer. An inter-island air-route complements the slower ferries. And improved transpor-tation has led to an increased flow of tourist t r a f f i c and a greater contact with the outside. Orkney i s playing a substantial role i n a twentieth century world. 2. Positive reciprocal stereotype script Their experience with the Hudson's Bay Company has l e f t Orcadians "export minded to an astonishing degree. The world is the Orcadians' familiar market place"" and Canada s t i l l proves to be the major importer of Orkney produce. Whether i t be whiskey, smoked salmon, or highland tweed, the great bulk of exports head toward Halifax or the St. Lawrence Seaway. The Orkney Islanders have been dealing with Canadians for over two hundred years now, and so they have developed a strong concept of them. This concept continues to be reinforced by recent Orcadian immigrants to Canada and by the descendents of former immigrants, who, although they have been wooed away from home, hate to lose contact with their native islands. 195 In general, the Orkney Islanders see Canadians as good-tempered individuals who have always behaved politely and broad-mindedly toward them in the past. This good-tempered manner and bright, respectful broad-mindedness remains a part of the Orcadian image of Canadians to this day. Commercial ties with Canada centre around importing frozen Atlantic salmon, smoking i t , and returning i t together with large consignments of tweed and moderate amounts of whiskey. If asked about their dealings with Canada, the islanders w i l l extend their definition of Canadians and describe them as reasonable, amusing, clean-cut, humourous, energetic, and interesting. If one finds i t odd that Canadians are so clearly defined i n the Orcadians' view of the world, i t i s well to remember that in over a thousand years of recent history the Orkney Islanders' relative isolation has been interrupted only by the regular v i s i t s of these Canadian, Hudson's Bay adventurers, who did much to add interest to the lives of people whose contact with Northern Scotland and Western Norway had grown stale with familiarity. The Hudson's Bay Company gave employment to out-of-work farmers, to enterprising businessmen, and to fishermen who could not survive on the meagre earnings of the sea. It i s not surprising, then, that the editorials in the "Orcadian" newspaper were f u l l of glowing accounts of the people "whom Orcadians considered to be their fellow Canadians." Words such as friendly, wise, reasonable, tolerant, humourous, appreciative, kind, generous, respectful, 196 sincere, co-operative, and a multitude of other flattering attributes, continue to dot the pages of newsprint. It is not a mistaken impression on the part of the Orkney Islanders to assume that there was a close tie between the developing Canadian nation and the historical and ancient Orkney Islands. Much of the cargo taken to the Bay came from Stromness Chandlers, who, perhaps more than any other group on the islands, started the export trade which s t i l l flourishes today. Furthermore, the company favoured Orcadians for service in Northern Canada, "At one stage nearly three-quarters of the Company's men were Orcadians. Of these—no fewer than ten Orcadians became governors or chief factors, and another eighteen d i s t r i c t masters or chief traders. About 1750 the Company's agent in Stromness was paying out up to fe3,000 in wages, annually 0." Even today, advertisements s t i l l appear in the Orcadian newspaper for young Orcadians to go to Canada in the Hudson's Bay service. No wonder that the fishing fleet has remained small and underdeveloped, and l i t t l e wonder, too, that many of the outlying islands have become depopulated and uninhabited. Yet, despite these upheavals i n the island l i f e , these changes were seen to have many benefits and their instigators seen as a people beyond reproach. The l i s t of favourable words used to describe them i s endless: clean-cut, appreciative, sincere, interesting, polite^and tolerant people who carry out business co-operatively and appreciatively, in a friendly, bright, energetic, and generous manner. 197 Indeed, many Orcadians demonstate spontaneous appreciation for the forgiving way i n which Canadians have handled lapses in business commitments. This forgiving quality, perhaps more than any other inferred characteristic, permitted the Orkney Islanders to see Canadians as being, above a l l , amusing, kind, and wise people. 3. Negative reciprocal stereotype script Their experience with the Hudson's Bay Company has l e f t Orcadians aware of the value of exporting their produce. "The world i s the Orcadians' familiar market place 1 1" and Canada s t i l l proves to be the major importer of Orkney products. whether i t be whiskey, smoked salmon, or highland tweed, the great bulk of exports head toward Halifax or the St. Lawrence Seaway. Business has been carried on between Canada and Orkney for over two hundred years now. In that time the island people as a whole have devel-oped a strong concept of Canadians. It i s a concept which continues to be reinforced by Orcadians who have been forced to immigrate through lack of work opportunities, and by the descendants of former immigrants who, although they have had to leave home, hate to lose contact with their native islands. In general, the Orkney Islanders see Canadians as ill-tempered individuals who have always behaved impolitely and narrow-mindedly toward them in the past. This ill-tempered manner and d u l l , disrespectful, narrow-mindedness remains a part of the Orcadian image of Canadians to this day. 198 Present commercial ties with Canada centre around importing frozen Atlantic salmon, smoking i t , and returning it together with large consignments of tweed and moderate amounts of whiskey. If asked about their dealings with Canadians, the Islanders w i l l readily extend their definition of Canadians and describe them as unreasonable, tiresome, lazy, and uninteresting. If one finds i t odd that Canadians are so clearly defined in the Orcadians view of the world, i t is well to remember that in over a thousand years of recent history, the Orkney Islanders have been caught at the mercy of opposing p o l i t i c a l forces, and the regular recruiting v i s i t s of the Hudson's Bay Company were, to many, the fi n a l straws in a long series of indignities. The Hudson's Bay Company offered employment to many able-bodied men. But in doing so i t stripped a poor and underdeveloped land of Its much needed labour force. It i s not surprising, then, that the editorials in the "Orcadian" newspaper were f u l l of unfavourable accounts of Canadians whom Orcadians considered to be the latest exploiters of their population. Words such as: unfriendly, foolish, unreasonable, intolerant, humourless, unappreciative, unkind, ungenerous, disrespectful, insincere, unco-operative, and a multitude of other unflattering attributes continue to dot the pages of newsprint. The present negative impression that the Orcadians have of Canadians has festered throughout the period in history when the Orkney Islands were connected with the development of the Canadian colonies. 199 Much of the cargo taken to the Bay came from Stromness Chandlers, who, perhaps more than any other group on the islands, started the export trade which s t i l l flourishes today. Indeed, the Hudson's Bay Company favoured Orcadians for service in northern Canada. It found that they were prepared to work for longer hours with less pay than their Scottish and English counterparts. At one stage nearly three-quarters of the Company's men were Orcadians, and around about 1750 eighty per cent of the males in Stromness had l e f t for overseas service. Unfortunately, however, the men were hired as labourers, or at best as craftsmen, and despite their large enrollment in the Company few were promoted or received positions of authority. So, while the fishing fleet remained underdeveloped and while much of the rich agricultural land lay fallow, the cream of Orkney youth and manhood was being recruited overseas to escape poverty. But in their escape they l e f t behind depopulated islands and shattered communities. Canadians were blamed for these upheavals in island l i f e and the l i s t of unfavourable words used to describe them i s s t i l l endless: unappre-ciative, insincere, impolite, intolerant, and even messy and uninteresting were and are among the most noteworthy ascriptions attributed to Canadians. They were seen by a frustrated people to be businessmen who carry out trade unco-operatively and unappreciatively, in an unfriendly, d u l l , lazy, and ungenerous manner. Indeed, many Orcadians demonstate spontaneous anger at the unforgiving way in which they think Canadians have handled understandable lapses i n 200 business commitments. This unforgiving quality, perhaps more than any other inferred characteristic, permitted the Orkney Islanders to see Canadians as being, above a l l , tiresome, unkind, and foolish people. Concluding remarks (used in a l l three scripts) These unique islanders, with their ancient indeterminate historical roots and their strong views about the world and the people i n i t , live f u l l l i v e s . They work long hours but they also find time to relax and enjoy themselves, whether i t be at a club meeting, a local pub, or the modern and recently built dance h a l l that has made electric music part of the Orkney l i f e today. Orkney is in a time of change and flux. But then i t has always been so. One has the feeling that these proud, self-possessed people with strong attitudes towards the world can cope with any change and new development. As a group they have been labelled with many characteristics. They can be described variously as meditative, inoffensive, moderate, although aggressive in business, discriminating, eccentric, ungraceful, c r i t i c a l , naive, conventional, and even irreligious. In many different ways each and a l l of these traits apply to the islanders. They are a complex people and they look to each new tomorrow with hope and expectation. * * * a. Bailey (1971) b. Bank Ahead (1967) c. Bank Ahead (1967) d. Bank Ahead (1967) e. Bailey (1971, p. 185) f. Bank Ahead (1967) g. Bailey (1971) h. Bank Ahead (1967) i . Bailey (1971) j . Bank Ahead (1967) k. Bank Ahead (1967) 1. Bank Ahead (1967) m. Bank Ahead (1967) n. Bank Ahead (1967) o. Bailey (1971) APPENDIX B: PROCEDURE FOR SELECTING THE TRAIT TERMS USED IN THE STIMULUS MATERIALS 203 1. To ensure that the documentary characterization of the Orkney Islanders was indeed neutral i t was decided to select the twenty t r a i t terms from the standard l i s t published by Anderson (1968). Procedure Five judges, three male graduate students and two working women, were given copies of both the Collier's Encyclopaedia (1972) and Encyclopaedia Britannica (1968) essays on the Orkney Islands and asked to: Please read through both of these essays about the Orkney Islands, paying particular attention to the descriptions of the history, economic organization, and population characteristics of the Islanders. After they had finished reading the brief passages they were presented with the one hundred and nineteen items from Anderson's (1968) l i s t that f a l l within the central range (375-225) of the 601-point "likableness dimension" (0-600), and given the following instructions: I would like you to consider each of the words in the following l i s t (starting with #203 and finishing with #321) and strike a line through those that would seem to you to be appropriate, logical, and/or meaningful in any description, however accurate or inaccurate, and however positive, negative, or neu-tr a l of the Orkney Islanders. Those words that were not deemed by any of the judges to be appro-priate Orcadian descriptors were deleted from the l i s t . The remaining items were divided into two sets, one containing words with likableness scores ranging from 301-375 and the other including terms scoring from 225-300. The words in each of these t r a i t pools were ranked according to the number of judges who had considered them to be suitable "attributes of the 204 Islanders." The ten most frequently chosen words in each of the subsets were selected to make up the f i n a l l i s t of twenty traits that formed the experimental characterization of the Orkney Islanders. Ties for tenth position In the rank order were resolved by a random draw and a check was made to determine that none of the terms were highly similar to each other in meaning. Results The twenty terms are listed in Table 1 in the main text. Their average likableness rating is 302.3. 205 2. The twenty positive and twenty negative traits that were used to represent the Orkney Islanders' characterizations of Canadians were also chosen from the Anderson (1968) l i s t . Procedure The items that f a l l within the upper limits (475-600) of Anderson's "likableness dimension" were compared with those words which have the lowest favourableness ratings (i.e., scores from 0-175) and a l l possible pairs of bipolar antonyms were formed. These forty-four antonym pairs were presented to five judges (one working woman, and one female and three male graduate students) who classified each pair with respect to whether or not i t s components could be considered meaningful in any possible definition of Canadians. The sets of antonyms were rank ordered in terms of the frequency with which they were considered to be potentially appropriate "Canadian attributes." The twenty most highly ranked, semantically non-overlapping word pairs were chosen to be included in the stimulus materials. The favourable traits formed the "positive reciprocal stereotype," while the negative terms were said to reflect the unfavourable Orcadian perspective of Canadians. Results The items are listed in Table 1 i n the main text. The average likableness rating of the set of positive terms is 492.4, while that of the negative traits is 102.4. 206 Note: The New American Roget's College Thesaurus in Dictionary Form (1962) and Funk and Wagnalls Standard Handbook of Synonyms, Antonyms, and Prepositions (1947) were the reference sources for both sections of this preliminary portion of the investigation. 207 APPENDIX C: PROCEDURE FOR SELECTING THE ADDITIONAL TARGET GROUPS 1 208 The choice of additional target groups was based on two considerations. From a methodological perspective i t was necessary to include a neutral and a negative target to provide a balanced set of stimuli, and from a procedural viewpoint i t was important to employ groups that the respondents were familiar with and that would serve to lend credence to the cover story. An empirical procedure, similar to the free response stereotyping format discussed in Chapter 2, was used to select two target groups which conformed to these requirements. Method Respondents. The participants in the study were forty-four undergraduates enrolled in 200- and 300-level psychology courses at the University of British Columbia. Twenty-three were male and twenty-one were female. They ranged in age from 18 to 34 with an average age of 21.3 and a modal age of 20. Procedure. Questionnaires were administered to volunteer respondents during free classroom periods. Having been instructed to provide personal information such as age and nationality the respondents were asked to: . . . write down the names of a l l of the ethnic, regional, and national groups that you can think of, regardless of whether or not you know a lot about their members, and regardless of whether or not you like the group in question. When they had completed this part of the task they were requested further to: 209 . . . read back through your l i s t of words and rate each of the groups on the following scale: unfavourable 1 2 3 4 5 6 7 8 9 favourable I L ! I J I I I I — o n the basis of your subjective impressions of the group members. Scoring procedure. A preliminary examination of the respondents' l i s t s was made and target names, like "cops" and "physicians," that were not associated with ethnic, regional, or national groups, were deleted. Salience weights, based on the order of occurrence of the responses, were applied to the remaining names within individuals' l i s t s . The weights were derived from the work of Szalay and his colleagues (cf. Szalay and Maday, 1973). Beginning with the f i r s t item they were applied in the following sequence: 6 , 5 , 4 , 3 , 3 , 3 , 3 , 2 , 2 , 1 , 1 Results. The responses were tabulated, the salience weights associated with each particular name were totalled, and the favourableness ratings were averaged. The group names were categorized into three sets on the basis of their average favourableness scores. Those with means ranging from 6.4 to 9.0 were classified as positively perceived groups. 'Neutral' groups were defined as those with scores in the 3.7 to 6.3 range, while those targets with highly unfavourable ratings (1.0 to 3.6) were considered to be negatively perceived outgroups. In a l l , nineteen positive, twenty-nine neutral, and three negative groups were listed by the respondents. The names within each set were ranked according to their total salience scores. The ten "most salient" names within each category are list e d in Table A. 210 Table A. The salience and favourableness scores of the most salient group names within each of the; three favourableness categories. POSITIVE GROUPS NEUTRAL GROUPS NEGATIVE GROUPS <u 6 rt I 91 0> T-i U 0> x> o o o> rt o T H O U : rt 3 rt-H co u o to 4-lrH-H 0) > CO O « 0 > > « 0 > 0) 0) o iH CU.C CO-H 00 Ur-I-H O « 0) I 0) 01 rH (-C X> o o)rt u bOMcn rtS l-i O CO o) > co > rt oi <(n c « 01 o rH 0143 CtJ>H bO 4J rH-H o rt o» I 0)0) rH M 43 O cu cou « 3 U OCD 0) >co > rt o) British Canadians Japanese French Canadians Italians Scots Ukranians British Columbians Western Canadians Maritimers 98 67 51 33 29 27 22 21 21 15 6.6 7.0 6.5 6.9 6.4 6.9 6.4 7.0 7.5 6.8 Chinese Americans Native Indians Irish Blacks French Poles Jews Germans Eastern Canadians 99 95 81 47 46 40 38 37 36 29 5.7 5.0 5.4 6.3 5.6 6.2 5.6 6.2 5.5 6.0 East Indians Arabs Pakistanis 85 3.6 17 2.1 6 2.5 211 Discussion. The findings reinforce the assumption that young students view Canadians in general as a salient as well as a positive reference group. The most salient neutral groups provided a number of possible choices for the neutral target stimulus. The European groups seemed to be the most appropriate of these and i t was decided to employ the Germans as an addi-tional target in the study because their favourableness score was quite close to the mid-point of the scale. Similarly, although East Indians were a more salient negative group i t was decided to use the Arabs as a target stimulus since they were more clearly perceived in unfavourable terms. (An example of the questionnaire is presented on the following two pages.) Page 1 212 PILOT STUDY I would be grateful i f you could provide information on the following variables: AGE: FEMALE OR MALE: NATIONALITY: Your anonymity, of course, i s guaranteed at a l l times. As we move through our everyday environment we encounter many different groups of people, who, each i n their own way, have special relevance to our lives. Academics, doctors, police persons and business people, among others, spring readily to mind. Although not quite so central to our cognitive maps of the social world, we are also familiar with a number of regional and national groups. For example, we often define ourselves as being Canadians, or Irishmen, or British Columbians. In a similar manner, we sometimes describe others as being Maritimers, Europeans, or Celts. To provide an indication of how Canadians categorize their social environment, I would like you to write down the names of a l l of the ethnic, regional and national groups that you can think of, regardless of whether or not you know a lot about their members, and regardless of whe-ther or not you like the group in question. Please turn the page and write your responses in those columns headed with an "A." Page 2 213 A B A B A B After you have written the names, read back through your l i s t of words and rate each of the groups on the following scale: Unfavourable 1 2 3 4 5 6 7 8 9 Favourable 1 I I I —on the basis of your subjective impressions of the group members. For example, low ratings indicate degrees of dislike and high ratings reflect relativeness attractiveness. Do not spend too much time on your ratings but do check to see that you have rated a l l of the words, putting your scores i n the appropriate spaces in columns headed with a "B." Thank you. APPENDIX D: POST-FILM QUESTIONNAIRE 215 Soc. Sciences/Educ. Film Proj. Drft.2 CODE NUMBER: Before you begin your assessments please make sure that you have completed the information section on the previous page. Remember, once again, to put your code number in the appropriate space on each page. Work through the questions rapidly, but always provide the most complete  information possible. Where possible, c i r c l e one of the options as the answer to the question under consideration. For example: Do you like folk music? (^T) NO DON'T KNOW Do you think that the film i s interesting? YES NO DON'T KNOW Do you think that the sound track should be changed? YES NO DON'T KNOW If "YES," in what way? Do you think that the visual sequences should be changed? YES NO If "YES," in what way? DON'T KNOW Do you think that a film of this sort should be: LONGER SHORTER THE SAME LENGTH If you answered "LONGER" or "SHORTER," specify the length that you think the film should be . Why? Do you think that the film i s relevant to Canadians? YES NO UNDECIDED Does i t say anything about Canadians? YES NO UNABLE TO SAY Do you think that you know more about the Orkney Islanders after watching this film? YES NO UNDECIDED If you answered "YES," how much more do you know about them? A LITTLE A MODERATE AMOUNT A LOT Has this new information changed your attitudes toward the Orkney Islands? YES NO UNABLE TO SAY If you answered "YES," in what way has your attitude changed? Be specific. 216 Write a couple of sentences on how you would like to see the film changed, on what you would like included and on what you would like dropped. Do you think that the l i f e s t y l e of the Orkney Islanders is DIFFERENT or THE SAME AS that of some Canadians? In what way is i t different or the same? Does i t appeal to you? YES NO If "YES," how much? A LOT SOME A LITTLE Do you think that the values of -the Orkney Islanders are THE SAME AS or DIFFERENT than that of some Canadians? In what way are they different or the same? Do you agree with their general values? YES NO DON'T KNOW If "YES," how much? A LOT SOME A LITTLE Do you think that the Orkney Islands are economically DEVELOPED or UNDERDEVELOPED? Is i t a RURAL or an URBAN society? Do you think that i t i s POORER THAN, RICHER THAN or the SAME as Canada, economically? How favourably do you think the islanders were described i n the film? VERY UNFAVOURABLY UNFAVOURABLY NEUTRALLY FAVOURABLY VERY FAVOURABLY How favourably do you view the islanders? VERY UNFAVOURABLY UNFAVOURABLY NEUTRALLY FAVOURABLY VERY FAVOURABLY Would you: lik e to v i s i t there, lik e to liv e there for a short time, live there for a long time, or couldn't care less? Do you think that the Orkney Islanders see other people VERY UNFAVOURABLY UNFAVOURABLY NEUTRALLY FAVOURABLY VERY FAVOURABLY Elaborate in a sentence or two. 217 Do you think that the Orkney Islanders see Canadians: VERY UNFAVOURABLY UNFAVOURABLY NEUTRALLY FAVOURABLY VERY FAVOURABLY Does this affect your attitude toward them? YES NO DON'T KNOW If so, how and in what way? Please be specific. Do you think that the film was worth watching? YES NO DON'T KNOW Do you think that films are useful educational devices? YES NO Explain. DON'T KNOW Please comment on any topic that you feel was not covered by these questions. Suggest improvements for the script i n terms of information, style, musical background, etc. Also suggest how a greater amount of Canadian relevance can be built into the script. Thank you for your co-operation. See you next week. APPENDIX E: EXAMPLE OF THE OPEN-ENDED STEREOTYPING FORMAT 219 Psych. Dept./Can. Viewpoints Pro j . M/l;' CODE NUMBER: Write down a complete l i s t of those traits and descriptive words that you  think are necessary to characterize the ORKNEY ISLANDERS adequately. Use as many spaces as you need in the columns marked with an A, and please write your words legibly at a l l times. Leave the smaller columns, that are marked with a B_, empty unt i l you have completed your l i s t of attributes and have read the instructions at the bottom of the page. ORKNEY ISLANDERS A B A B_ A B A B Read back over the words that you have used to describe the ORKNEY ISLANDERS and rate each one on the following scale: Unfavourable 1 2 3 4 5 6 7 8 9 Favourable t i ! i i i t t > — o n the basis of the desirability of these- attributes in friends and acquaintances. For example, i f you gave a particular t r a i t a rating of 3, 2, or 1, i t would indicate that you think that the t r a i t i s unfavourable, and i s , therefore, undesirable in friends and people you know. "Vain" and "grouchy" are two such characteristics. On the other hand, the attributes that you would give high ratings to are those that you consider to be favourable and desirable in people. Put your scores opposite the words in the appropriate spaces in those columns that are headed with a B_. Do not spend too much time on your ratings, but do check to see that you have rated a l l of the words that  you have l i s t e d . APPENDIX F: EXAMPLE OF THE CHECK LIST MEASURING PROCEDURE Psych.Dept./Can. Viewpoints M/l CODE NUMBERS: 221 1. Read carefully through the l i s t of tra i t words and tick (•) those which seem to you to be typical of the ORKNEY ISLANDERS. If you do not find appropriate words for a l l of the typical ORKNEY ISLANDER character-i s t i c s you may WRITE IN those which you think are necessary to complete an adequate description. Tick) Star Rat-ing Tick/ Star Rat-ing Tick/ Star Rat-ing ILL-TEMPERED POLITE MODERATE SCIENTIFICALLY-MINDED UNINTERESTING PERSISTENT UNFRIENDLY IRRELIGIOUS UNREASONABLE AMUSING REASONABLE PROUD ENERGETIC HUMOURLESS UNFORGIVING SENTIMENTAL FEARLESS JOVIAL SELF-CENTRED DISRESPECTFUL CLEAN-CUT CO-OPERATIVE STRAIGHT FORWARD NAIVE NARROW-MINDED LAZY SENSUAL FOOLISH UNGRACEFUL BROAD-MINDED SINCERE MATERIALISTIC BRIGHT TIRESOME EFFICIENT PHYSICALLY DIRTY ' QUIET KIND PLEASURE-LOVING GENEROUS DISCRIMINATING INTOLERANT cont... 222 TOLERANT MEDITATIVE INOFFENSIVE INSINCERE GOOD-TEMPERED UNAPPRECIATIV 5 SELF-POSSESSED MESSY FRIENDLY TREACHEROUS STOLID WISE DAYDREAMER UNKIND AMBITIOUS HUMOUROUS AGGRESSIVE CRUEL ECCENTRIC APPRECIATIVE CONVENTIONAL FORGIVING RESPECTFUL METHODICAL INTERESTING IMPOLITE EXTREMELY-NATIONALISTIC CRITICAL DULL DECEITFUL MEAN METICULOUS BLUNT 2. Go back over the l i s t of words and mark, with a star (*), the TEN traits which seem to you to be MOST typical of the ORKNEY ISLANDERS. Be sure to star TEN words, even i f the original l i s t of ticked words happens to be shorter. 3. Please read once more over the l i s t of traits that you have ticked and starred and rate each of the words on the same scale that you used in the previous booklet: Unfavourable 1 2 3 4 5 6 7 8 9 Favourable l ! ! I I I t I I 4. Please remember to check to see: (1) that you have rated a l l of the words that are ticked (and/or starred), and (2) that you have starred (*) the TEN most typical characteristics of ORKNEY ISLANDERS. WRITE DOWN ADDITIONAL WORDS HERE IF YOU FEEL THAT THEY ARE NEEDED TO COMPLETE THE FULL DESCRIPTION: The original version of the check l i s t forms fitted on a single foolscap page. 224 APPENDIX G: PERSONAL INFORMATION QUESTIONNAIRES USED IN STEP 1 AND STEP 2 OF THE EXPERIMENT 225 Soc.Sciences/Educ.Film Proj. Drft.2 CODE NUMBER: There is a card with your code number on i t attached to the top of this page. It guarantees your anonymity in both this and the next study i n which you w i l l be participating. The answers to the questions w i l l be analyzed with reference to this code number and not to your name. Because you w i l l be using the same code number again next week, i t is important to keep the card and bring i t with you next time. Please write your code number on the top right hand corner of this page and write i t in the appropriate space on each of the following pages. PERSONAL INFORMATION: Age: Date of Birth:  Sex: Male / / Female / / Occupation:  Nationality: Canadian / / Other: (say which)  Birthplace:  Location of Permanent Residence: B.C. / / Elsewhere: (say where) If you are not a full-time student: 1. How many courses have you taken since leaving school? 2. What were they? Do you intend to take courses in the future? Yes / / No / / If so, in what subject? Have you ever lived abroad? Yes / / No / / If so, where? For how long? Thank you for your co-operation. The material in this section and your replies to the questionnaires w i l l be kept confidential at a l l times. Likewise, the project staff would appreciate i t i f you would refrain from discussing the details of the film script with your friends and colleagues. Psych.Dept./Can. Viewpoints M/l CODE NUMBER: 226 Please write your code number on the top right hand corner of this page and write i t in the appropriate space on each of the following pages. If, for some reason, you have been unable to bring your code number with you, please write down the colour of the card on which i t was printed as well as the time that you were tested on last week. This should help us in retracing the missing number (by a process of elimination). Also, please contact the study supervisor about the matter. Before beginning the task we would like you to provide some personal information. The material in this section and your replies to the ques-tionnaires w i l l be kept confidential at a l l times. Your anonymity is assured throughout the study. PERSONAL INFORMATION: Age: Date of Birth: Sex: Male / / Female / / Occupation:  Nationality: Canadian / / Other: (say which)  Location of Permanent Residence: B.C. / / Other: (say where) Have you ever lived abroad? Yes / / No / / If so, where? For how long? Thank you for your co-operation. We would appreciate i t i f you would pay close attention to the instructions in the following tasks and carry them out as best you can. When something seems unclear to you don't hesitate to ask a question about i t . APPENDIX H: TABULATION SHEET FOR NOTE-TAKING DURING FILM SEQUENCE 228 Soc.Sciences/Educ. Film Proj. Drft.2 Before we show the film, we would like to say that i n previewing the script that accompanies the film we f e l t that the narration was a bit too rapid, and that there was, perhaps, too much information on the soundtrack. We hope that this w i l l not affect your concentration on the script too much. What we would like you to do is to make notes on how much information needs to be l e f t out, in your opinion. Also make note of which information should be l e f t out in the revised edition. NOTES What percentage of the film script should be excluded in later editions? What sections should be l e f t out? Be as specific as possible. APPENDIX I: "SOURCES OF STEREOTYPE INFORMATION" QUESTIONNAIRE 230 Psych.Dept./Can. Viewpoints Proj. M/l CODE NUMBER: In addition to measuring your views on the four groups of people i t i s important to find out the things that influenced you in forming your ideas. Please l i s t a l l of the major sources of information (e.g., books, films, conversations, and so forth) that you have looked to when forming your impressions of each of the groups. Be very specific (e.g., Hawaii Five-0, Market Place, Sun Editorials, This Country in the Morning) where possible. ARABS: CANADIANS: GERMANS: ORKNEY ISLANDERS: Write two or three sentences on how you think your ideas about different social, regional and national groups are formed: APPENDIX J: BASIC RIGHTS AND PRIVILEGES OF VOLUNTEER SUBJECTS 232 DEPARTMENT OF PSYCHOLOGY THE UNIVERSITY OF BRITISH COLUMBIA Basic Rights and Privileges of Volunteer Subjects Any person who volunteers to participate in experiments conducted by f u l l or part-time members of the faculty of the Department of Psychology at the University of British Columbia, by their employees, or by the graduate and undergraduate students working under the direction of faculty members of the above named Department, i s entitled to the following rights and privileges. 1. The subject may terminate and withdraw from the experiment at any time without being accountable for the reasons for such an action. 2. The subject shall be informed, prior to the beginning of an experiment of the maximum length of time the experiment might take and of the general nature of the experiment. 3. The subject shall be informed, prior to the beginning of an experiment, of the nature and function of any mechanical and ele c t r i c a l equipment which is to be used in the experiment. In cases where the subject i s in direct contact with such equipment, he shall be informed of the safety measures designed to protect him from physical injury, regard-less of how slight the possibility of such injury i s . 4. The subject shall be informed, prior to the beginning of an experiment, of the aspects of his behavior that are to be observed and recorded and how this i s to be done. 5. Any behavioral record that is obtained during the course of the experiment i s confidential. Any behavioral records that are made public through either journal papers or books, public addresses, research colloquia, or classroom presentations for teaching purposes, shall be anonymous. 6. The subject shall be offered, at the end of an experiment, a complete explanation of the purpose of the experiment, either orally by the experimenter or, at the option of the experimenter, in writing. The subject shall also have the opportunity to ask questions pertaining to the experiment and shall be entitled to have these questions answered. 7. The subject has the right to inform the Chairman of the Departmental Committee of Research with Human Subjects of any perceived violations of, or questions about, the aforementioned rights and privileges. APPENDIX K: VERBAL DIRECTIONS TO RESPONDENTS BEFORE AND AFTER FILM PRESENTATION i 234 Pre-film directions " I l l l go over the handout again just to make sure that a l l of the details are clear. "Please ask questions i f you want anything explained or i f there are some points that you want to know more about. "As you may know, U.B.C., like most universities, has a supply of films that can be used as teaching aids. While such films should be interesting and informative, most of the films that are now available are either too long, too detailed, or too irrelevant to serve the needs of the various courses. This problem is particularly evident in adult- edUcation courses. "There is a growing body of professors who are very concerned about this matter. "A number of teachers have suggested that the psychology, sociology, and anthropology departments, among others, should start their own film libraries in order to f i l l some of the existing demands. Indeed, a lot of enthusiasm has centred around this proposal. "As a result this pilot project was established to put together a couple of t r i a l films to see i f the general idea of each department doing 'its own thing' would work. "It was decided that we would make a beginning by working on a couple of short, all-purpose, documentary-type films. "Our aim is to give something of the flavour of the various lif e s t y l e s in different types of communities and cultures around the world. 235 "We hope to describe the characteristics of various groups of people; to portray their lifestyles and means of earning a l i v i n g ; to say something  about how they view other groups in the world around them; in general, to Outline their attitudes, customs, and habits. "To start with we are searching out ideas and getting down to work on a number of different scripts for a film about the ORKNEY ISLANDERS. We thought that this group was particularly interesting because i t s l i f e s t y l e has been shaped around i t s relative isolation, and although few people know very much about the Islanders they played a central role in early Canadian history and the development of the Hudson's Bay Company. "In an effort to make the finished documentaries as interesting as possible we are trying to involve a wide range of people in their develop-ment. Consequently we are asking a number of groups, including you, to assess and suggest improvements on variations of particular scripts. By acting on your advice, as well as that of the social science course instructors, we hope that we can blend and select from among the alterna-tives to produce the f i r s t of a number of useful films. "The film that you are about to see i s original footage. The subtitles w i l l give you an indication of where and when i t was made. "The different scenes seem to say a lot about l i f e on the Orkney Islands, so we were very reluctant to change them. What we have done, however, is to completely alter the soundtrack and add new descriptions and information. "Because this is the f i r s t draft of a tape-recorded soundtrack, the synchronization w i l l be somewhat crude. 236 "We must warn you not to be disturbed by unheard conversations where peoples' li p s move while you cannot hear what they are saying. Please remember that the fi n a l product w i l l be made by a professional company and w i l l be much more polished. As yet this i s only the spade work! (Momentary pause) (Emphasis) "Rather than attend to the production aspects of the film, i t is better i f you concentrate on the information contained in the scenes and in the soundtrack. "Try to form a general impression of the visual scenes and pay close  attention to the soundtrack. This w i l l help you in your assessment of the film. "Look for the various descriptions of the general character of the Islanders. Note their principal occupations and the types of links they have with the outside world. "Also attend to the information that may give you some idea of how they view various 'outsiders,' thus getting a general idea about their attitude to the world and the people in i t . "Decide whether this has any relevance to you as a Canadian. "After I've shown the film I ' l l hand each of you an evaluation booklet to be completed. "But before we start I'd like to say one more thing. When we pre-viewed the film that you are just going to see we f e l t that the narration was a bit too rapid, and that there was perhaps, too much information on 237 the soundtrack. We hope that this w i l l not affect your concentration on the script too much. "... What we would like you to do i s to make notes on how much infor-mation needs to be l e f t out, in your opinion. I ' l l hand out copies of this form (Appendix H) for you to make notes on. Where possible you should be specific about which items of information should be l e f t out in the f i n a l edition." Post-film directions To ensure that the instructions were understood the opening paragraph in the booklet was paraphrased: "The code number on your coloured card guarantees your anonymity in the study. Please write i t at the top of each of the four pages in the booklet. "Since I hope that most i f not a l l of you can come back next week to take part in another study I would like you to keep the card u n t i l then. Remember i t s colour and try to remember the code number. "For the present I'd like you to f i l l out the information section on page 1. This w i l l help us when we analyze and interpret your replies to the various questions on the following pages. When you have done this turn the page and work rapidly through the questions, giving the most complete information possible." APPENDIX L: INTRODUCTORY PASSAGE FOR PHASE TWO OF PROCEDURE 239 Psych.Dept./Can. Viewpoints Proj. M/l Very often psychologists and sociologists have measured the personalities, customs,, and life s t y l e s of people without paying too much attention to the ways in which individuals actually view the world. As a result we know very l i t t l e about the differences between, for example, Canadians and Norwegians in the way they see the people and national groups which surround them. However, i t i s becoming increasingly evident that the understanding of peoples' perceptions of their social world is of the utmost importance. The present study has been established to discover the ways in which Canadians view a wide number of different social groups. In addition, we are also trying to determine how Canadians think other people look back at us, and at the world In general. Consequently, you w i l l be asked to describe various groups and to describe  the same four groups as you think you would do i f you had been born and  raised in another Culture. Because of i t s wide scope, the research i s being carried out step by step, with the focus being placed on different cultural groups on each separate occasion. Thus, the groups that you describe w i l l not be the same as those that other sets of participants have been asked to characterize. Some of the groups that you may be faced with w i l l be unfamiliar to you. You w i l l know quite a lot about others. In either case, please perform  as best you can on a l l of the tasks. APPENDIX M: VERBAL DIRECTIONS TO RESPONDENTS DURING PHASE TWO OF PROCEDURE 241 Step 1 "When we deal with individuals on an everyday basis or when we meet new people for the f i r s t time we form impressions of them, and we get ideas about what they like and dislike, about what types of traits they have—for example some people are friendly and others are pushy—and sometimes we may even be able to figure out how they think about us and about other people we know. "In the same way we form impressions about groups of people. I bet i f I asked any of you to describe the Officer Cadets or the "Two Weekies" you could do so without too much trouble. (There was generally a murmur of agreement at this point.) I bet, too, that a lot of you have a f a i r idea about what the Officer Cadets and other groups, lets say the Bandsmen, think of you as a group. (Laughter) "As a psychology student my main interest i s in this broad question of how we view individuals and groups in the world around us. "I'm particularly interested in the ideas that Canadians have about other national and regional groups. "As you can see from the handout, I am involved i n a large project which w i l l give us precise information not only about how Canadians view a wide number of different social groups but also about whether or not Canadians have ideas about how various nations see us and the world in general. "We are trying to work with as many different people as possible, f i r s t l y to lighten the work load on each particular set of individuals and, secondly, to give us as large a sample as possible so that we can be more certain about the accuracy of our findings. 242 "Actually, of a l l of the groups we have worked with so far you are the most ideal because you represent a wide section of the Canadian population—before now we have had trouble finding people who l i v e outside of B.C. "Your job this evening f a l l s into two parts. I ' l l explain the second part to you later on and we'll just deal with the f i r s t section for the moment. I ' l l hand around these booklets before we begin (Pause) "As you f l i p through the pages you'll see that you are being asked to describe four different groups in your own words. They are the Germans, Arabs, Canadians, and Orkney Islanders. "We are asking everybody who takes part in this project to characterize Canadians. The Germans and Arabs are being used i n this case because they are among the groups with which most people are familiar. The Orkney Islanders are list e d because as soon as my colleagues on this particular project heard that you had taken part in last week's "curriculum study" they f e l t that i t would be useful to obtain your views about the group. They also thought that i t would be worthwhile to ask you to put yourselves In the shoes of the Islanders and to give us your ideas about how they see themselves and each of the other three groups that are listed here. "The example on the second and third pages of the booklet should help you to become familiar with the instructions and should give you a clear idea about what needs to be done in the task. The target in the example is a f i c t i t i o u s one and the words used to describe i t are in another language so that they won't influence your descriptions of the four 243 targets in the study. In a l l other respects the example is the same as the questionnaires that follow the personal Information form. "I am going to work through one other example on this overhead projector. In doing so I w i l l try to give you a sense of what you should consider when choosing a word with which to describe a group." The experimenter then activated a portable overhead projector, focused the image of the plastic lay-on of a free response questionnaire, read the instructions aloud, and responded to each one in turn verbalizing his thoughts as he did so., (Instr.) "Write down a complete l i s t of those traits and descriptive words that you think are necessary to characterize FILMSTARS adequately. "Use as many spaces as you need in the columns marked with an A, and please write your words legibly at a l l times. Leave the smaller columns, that are marked with a B, empty unt i l you have completed your l i s t of attributes and have read the instructions at the bottom of the page. (Resp.) "Let's see. When I think of film stars I think of them as being "talented" so I ' l l print TALENTED in this wide column here (Column A, row 1). I also think of them as being "rich" so I ' l l put that down. By and large, too, I think of American film stars rather than those from other countries so I ' l l add AMERICANS to my l i s t . Some film stars are "tempermental" but I don't think a majority of them are. I ' l l leave out that word. However, I do think that a great many of them are "self-important" so I ' l l write that down ... (Pause) ... "I can't really think of any more words off-hand so I ' l l stop and see what I have to do next. 244 (Instr.) "Read back over the words that you have used to describe FILMSTARS and rate each one on the following scale: Unfavourable 1 2 3 4 5 6 7 8 9 Favourable i ! i i i i i t » —on the basis of the desirability of these attributes in friends and acquaintances. "For example, i f you gave a particular t r a i t a rating of 3, 2, or 1 i t would indicate that you think that the tr a i t is unfavourable, and i s , therefore, undesirable in friends and people you know. "Vain" and "grouchy" are two such characteristics. "On the other hand, the attributes that you would give high ratings to are those that you consider to be favourable and desirable In people. "Put the scores opposite the words, in the appropriate spaces in those columns that are headed with a B_. Do not spend too much time on your ratings, but do check to see that you have rated a l l of the words that you have l i s t e d . (Resp.) "O.K. Let's go back and make the ratings. The f i r s t word I list e d i s "talented." That i s a desirable characteristic, I ' l l rate i t 8. The next word i s "rich." It i s certainly not a bad attribute to have, but i t isn't great either. I ' l l give i t 6. It's on the positive side of the mid-point as far as my personal feelings are concerned. The next word is "Americans." From my viewpoint i t i s neither desirable nor undesirable to be American so that term gets a 5. The last word in my description i s "self-important." I think that i t i s an undesirable characteristic. I ' l l rate i t 2. (Pause) 245 "To recap, I have li s t e d a number of words that I think satisfactorily and adequately describe FILMSTARS. From my point of view, fewer words would have been an inadequate and incomplete description, while more words would have overdone the case. You w i l l know yourself when your descriptions meet the proper requirements. Essentially you have to be thorough without spending too much time or using too much effort on the task. It i s not necessary to strain the limits of your memory. Just write those words that readily come to mind. "It i s also very important to rate every word that you have lis t e d . Again, do not spend too much time pondering on your ratings but do check to see that a l l of the words have been given a value. Don't forget to put your code number on the top of every page and be sure to carry out a l l of the instructions as they are stated. (The overhead projector was switched off.) "Bring your booklet to me when you have finished the task. I ' l l check through i t to see that everything is in order and I ' l l explain the second part of the study to you." Step 2 "This booklet i s very similar to the last one. It contains two sections in addition to an example. Read the example carefully and carry out a l l of the instructions on a l l of the pages. The instructions are somewhat more detailed than they were in the last booklet, but they are not any more d i f f i c u l t . Looking through the pages you w i l l see that you are being asked again to characterize the Arabs, Canadians, Germans, and 246 Orkney Islanders, f i r s t from your own point of view and then from the point of view of an Orkney Islander. This time, however, you have been provided with a l i s t of 75 words to choose from when describing the groups. "If you turn to the example or to one of the pages you w i l l see the exact sequence of instructions that you have to follow. "1. Read carefully through the l i s t of tra i t words and tick (V) those which seem to you to be typical of (let us say) the ORKNEY ISLANDERS. If you do not find appropriate words for a l l of the typical ORKNEY ISLANDER characteristics you may WRITE-IN those which you think are necessary to complete an adequate description. "2. Go back over the l i s t of words and mark, with a star (*), the TEN traits which seem to you to be MOST typical of the ORKNEY ISLANDERS. Be sure to star TEN words, even i f the original l i s t of ticked words happens to be shorter*3. "3_. Please read once more over the l i s t of traits that you have ticked and starred and rate each of the words on the same scale that you used in the previous booklet: Unfavourable 1 2 3 4 5 6 7 8 9 Favourable t t t | l I I ! J "4_. Please remember to check to see: (1) that you have rated a l l of the words that are ticked (and/or starred), and (2) that you have starred (*) the TEN most typical characteristics of ORKNEY ISLANDERS. (Pause) "It i s essential that you carry out a l l of the instructions. If you work carefully and quickly i t should not take you long to complete the booklet. 247 "Work through each of the pages in order, finishing one before you start another. "If you have any comments we would welcome them, just write them on the back of the pages. "Before you go back to your seats write your code number on the top of every page and when you have finished bring your booklet to my desk so that I can check i t as I did your other one. Thank you." Notes a. Cadets who spend two weeks at the camp for a short introductory course. b. Eventually i t was decided to exclude the responses that had been starred but not ticked from the data analyses on the grounds that they were 'forced' rather than 'spontaneous' selections. The remaining responses that had been both starred and ticked were not treated differently than those which had merely been ticked. APPENDIX N: EXAMPLE OF CHECK LIST RESPONSES 249 Psych.Dept./Can. Viewpoints Proj. M/l CODE NUMBER: In the last two booklets you w i l l be asked to describe the same four groups, this time using a check l i s t from which you select suitable words. You w i l l be asked to tick ( •/) words which appropriately describe the group in question. You w i l l be asked, further to pick out and star (*) the ten most typical descriptive words. Finally, you w i l l have to rate a l l of the words that you have selected on the same nine-point scale that you used before. Again, we w i l l present you with an example which should i l l u s t r a t e the task. EXAMPLE 1. Read carefully through the l i s t of tr a i t words and tick (>/) those which seem to you to be typical of the FIRBOLGS. If you do not find appropriate words for a l l of the typical FIRBOLG characteristics, you may write i n those which you think are necessary to complete an adequate description. Tick, Star Rat-ing Tick, Star Rat-ing Tick Star Rat-ine OG * • 7 CASTA FANACHT BEAG CHEIM • 4 RAIBM FIONN * • 9 BMEITH * • 8 BLAISTIUL ATMAS * • 7 BLIANA GUIDEACH • 3 TEANGA * • 7 LEANTA * • 6 POBAL y 5 EISTEACMT • 5 SINGIL * • 4 CHUR POSTA AMADACH V 3 CEILE FILI • 5 ANUAS * • 6 TARLU 6 250 Psych.Dept./Can. Viewpoints Proj. M/l CODE NUMBER:  2. Go back over the l i s t of words and mark, with a star (*), the TEN traits which seem to you to be MOST typical of the FIRBOLGS. Be sure to star TEN words, even i f your original l i s t of ticked words happens to be shorter. 3. Please read once more over the l i s t of traits that you have ticked and starred and rate each of the words on the same scale that you used in the previous booklet. Unfavourable 1 2 3 4 5 6 7 8 9 Favourable t t i t i t i i i Note that I did not tick a l l of the words in the l i s t but only those that  I considered to be appropriate in describing the Firbolgs. Notice also, that I happened to tick 15 words. This meant that I had to pick out 10 of these 15 as being most typical of the Firbolgs. These were the ten traits that I starred. I starred exactly 10 t r a i t s — n o more and no less. If my l i s t had been shorter, say six or seven words, I would have starred a l l of these words as being most typical of the Firbolgs. I would also have chosen three or four more words which could be applied to the group (even i f not very well) and starred those too. In that way I would have ended up with 10 starred words. It i s important to remember, then, that however long or short your l i s t  of ticked words i s , you have to end up with ten starred words for each Look at the example once again to see what I have done. You w i l l observe that a l l of the words which have been ticked and/or starred have also been rated. This i s really important. Start on the tasks. Do not spend too much time pondering on your choices  but do check to see that a l l of the words have been rated and that exactly  ten of them have been starred. Don't forget to put your code number on the top of every page and be sure to carry out a l l of the instructions as they are written. 251 APPENDIX 0: TABLE OF RESPONSES TO QUESTIONS CONCERNING THE RESPONDENTS' POST-FILM IMPRESSIONS OF THE ORKNEY ISLANDERS Table A. The number and percentage of individuals i n each of the experimental groups and In the total sample choosing each of the possible responses to the questions concerning the respondents' post-film impressions of the Orkney Islanders Response Frequency Positive Non- Negative Total Question (paraphrased) Categories and % of reciprocal reciprocal reciprocal Sample response group group group (N=193) (n=58) (n=67) (n=68) RURAL n % 44 75.9% 46 68.7% 41 131 60.3% 67.9% Is i t a RURAL or an URBAN society? URBAN n % 10 17.2% 11 16.4% 17 38 25.0% 19.7% NO RESPONSE 4 6.9% 10 14.9% 10 24 14.7% 12.4% Table A (continued) Response Frequency Positive Non- Negative Total Question (paraphrased) Categories and % of reciprocal reciprocal reciprocal Sample response 3 group group group (N=193) (n=58) (n=67) (n=68) DEVELOPED n % 36 62.1% 30 44.8% 27 93 39.7% 48.2% Are the islands econom-ic a l l y DEVELOPED or UNDERDEVELOPED? UNDER-DEVELOPED NO RESPONSE n 20 34.5% 2 3.5% 31 46.3% 6 9.0% 31 82 45.6% 42.5% 10 14.7% 18 9.3% Table A (continued) Response 7Frequency Positive Non- Negative Total Question (paraphrased) Categories and % of reciprocal reciprocal reciprocal Sample response group group group (N=193) (n=58) (n=67) (n=68) RICHER n % 4 6.9% 2 3.0% 1 1.5% 7 3.6% Are the islanders RICHER, POORER, or THE SAMEXas Canadians? THE SAME POORER n % 14 24.1% 36 62.1% 7 10.5% 49 73.1% 12 17.7% 49 72.1% 33 17.1% 134 69.4% NO RESPONSE n % 6.9% 9 13.4% 6 8.8% 19 9.8% Table A (continued) Response Frequency Positive Non- Negative Total Question (paraphrased) Categories and % of reciprocal reciprocal reciprocal Sample response group group group (N=193) (n=58) (n=67) (n=68) Are the VALUES of the Islanders THE SAME as or DIFFERENT from those of Canadians? THE SAME DIFFERENT n % 18 31.0% 18 31.0% 13 19.4% 31 46.3% 13 44 19.1% 22.8% 23 72 33.8% 37.3% NO RESPONSE n % 22 37.9% 23 34.3% 32 77 47.1% 39.9% U i U i Table A (continued) Response Frequency Positive Non- Negative Total Question (paraphrased) Categories and % of reciprocal reciprocal reciprocal Sample response group group group (N=193) (n=58) (n=67) (n=68) Is the LIFESTYLE of the Islanders THE SAME as or DIFFERENT from that of Canadians? THE SAME DIFFERENT NO RESPONSE n n % n % 26 44.8% 22 37.9% 10 17.2% 20 29.9% 33 49.3% 14 20.9% 13 19.1% 45 66.2% 10 14.7% 59 30.6% 100 51.8% 34 17.6% Percentages are column percentages 257 APPENDIX P: TABLES OF MEANS AND STANDARD DEVIATIONS OF SCORES ON THE DIMENSIONS OF THE ESTIMATED RECIPROCAL ORKNEY ISLANDER STEREOTYPES OF CANADIANS 258 Table A. Means and standard deviations of the commonness and 'percentage of common terms' scores for the core and peripheral components of the respondents' estimates of the Orkney Islanders' reciprocal stereotypes of Canadians Stereotype Component Positive reciprocal group (n=58) Non-reciprocal group (n=67) Negative reciprocal group (n=68) co <U u o u CO 01 CO cu e c o o Core M Periphery M SD SD 1.8 4.7 1.2 2.4 1.6 4.4 1.1 2.2 1.2 3.7 0.9 1.8 e o O CO a cu u 14-1 O o o CO CU bO-tfl CO t e CU 0) O 4J U CU P. Core M g D Periphery M SD 36.1 31.1 24.8 13.3 3,1 ,Q 31:9 22.8 31.0 13.5 28.0 21.2 25.3 12.6 259 Table B. The mean numberrand percentage of positive, neutral, and negative words in the core and peripheral components of the respondents' estimates of the Orkney Islanders' stereotypes of Canadians Stereotype Favourableness Component of words Positive Non- Negative reciprocal reciprocal reciprocal group group group (n=58) (n=67) (n=68) FAVOURABLE M SD M% - SD 3.48 2.3 66.1% 34.4 3.33 2.9 60.8% 29.7 21.60 1.7 33.6% 32.5 Core NEUTRAL M SD *%SD 1.16 2.4 15.1% 20.6 1.22 1.5 21.8% 24.8 1.06 1.5 19.7% 24.9 M O T, - SD 0.78 1.2 0.94 1.4 1.96 1.6 UNFAVOURABLE ^SD 13.8% 23.9 17.4% 23.4 46.0% 37.1 FAVOURABLE M SB SD M% SD 10.55 7.4 64.2% 29.5 7.31 4.9 52.5% 30.1 5.02 4.8 33.4% 27.2 Periphery NEUTRAL M SD M% SD 3.90 4.1 25.1% 23.5 5.45 4.5 35.8% 24.8 5.27 8.3 30.6% 24.3 UNFAVOURABLE M SD M% SD 1.45 2.3 10.7% 15.9 1.79 2.9 11.7% 16.6 5.44 5.6 36.0% 32.7 260 Table C. Mean number of terms i n the core and peripheral components of the respondents' estimates of the Orkney Islanders' reciprocal stereotypes of Canadians that are common to the stimulus material characterizations of the Orkney Islanders and Canadians Positive Non- Negative Stereotype component reciprocal reciprocal reciprocal group (n=58) group (n«67) group (n=68) cu Xi CO c u o V •H T3 4J C ct) CO N i-H •H CO U H CU 4J O 0) CO C U M CO u Xi o o cu B .e. O Core M SD 0.2 0.4 0.1 0.3 0.1 0.4 Periphery M 3.1 3.0 3.5 SD 2.1 1.8 2.6 e Q O CJ C •H CO B cu u cu 1 c CO O e e rH O •H i-l 05 <4-( U c cfl CO 0) N t-t > *H T3 1-4 Vi CO 4J CU c •H 4J to CO o O o CO CX. u CO J3 o o B c rH O •H •H CO M-l 4-1 e CO CO CU N •H > •H "0 •H Vi CO 4J CU c CO 4-> CO 60 u u cu cfl c Vi CO Xi o Core Periphery Core Periphery M SD M SD M SD M SD 1.1 1.0 7.9 4.5 0.1 0.2 1.8 2.6 0.7 0.8 6.7 3.6 0.1 0.2 2.3 2.5 0.3 0.5 4.3 4.7 0.4 0.7 5.4 4.6 Note: Maximum score possible = 20 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0094149/manifest

Comment

Related Items